StevenLimcorn commited on
Commit
1db53a0
1 Parent(s): a214e6b

End of training

Browse files
all_results.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "eval_cer": 0.296368909038925,
4
+ "eval_loss": 1.1786432266235352,
5
+ "eval_runtime": 126.0181,
6
+ "eval_samples": 2958,
7
+ "eval_samples_per_second": 23.473,
8
+ "eval_steps_per_second": 2.936,
9
+ "eval_wer": 0.8593644354293442,
10
+ "train_loss": 5.91964619928868,
11
+ "train_runtime": 36587.1788,
12
+ "train_samples": 6380,
13
+ "train_samples_per_second": 17.438,
14
+ "train_steps_per_second": 0.544
15
+ }
eval_results.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "eval_cer": 0.296368909038925,
4
+ "eval_loss": 1.1786432266235352,
5
+ "eval_runtime": 126.0181,
6
+ "eval_samples": 2958,
7
+ "eval_samples_per_second": 23.473,
8
+ "eval_steps_per_second": 2.936,
9
+ "eval_wer": 0.8593644354293442
10
+ }
nohup.out CHANGED
@@ -15707,3 +15707,26 @@ Configuration saved in ./preprocessor_config.json
15707
 
15708
  Dropping the following result as it does not have all the necessary fields:
15709
  {'dataset': {'name': 'common_voice', 'type': 'common_voice', 'args': 'zh-TW'}}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
15710
  0%| | 0/370 [00:00<?, ?it/s]
15711
  1%| | 2/370 [00:00<00:43, 8.45it/s]
15712
  1%| | 3/370 [00:00<01:04, 5.69it/s]
15713
  1%| | 4/370 [00:00<01:15, 4.82it/s]
15714
  1%|▏ | 5/370 [00:01<01:26, 4.23it/s]
15715
  2%|▏ | 6/370 [00:01<01:28, 4.13it/s]
15716
  2%|▏ | 7/370 [00:01<01:32, 3.93it/s]
15717
  2%|▏ | 8/370 [00:01<01:36, 3.76it/s]
15718
  2%|▏ | 9/370 [00:02<01:32, 3.92it/s]
15719
  3%|▎ | 10/370 [00:02<01:33, 3.84it/s]
15720
  3%|▎ | 11/370 [00:02<01:29, 4.00it/s]
15721
  3%|▎ | 12/370 [00:02<01:27, 4.09it/s]
15722
  4%|▎ | 13/370 [00:03<01:25, 4.16it/s]
15723
  4%|▍ | 14/370 [00:03<01:28, 4.02it/s]
15724
  4%|▍ | 15/370 [00:03<01:22, 4.33it/s]
15725
  4%|▍ | 16/370 [00:03<01:23, 4.22it/s]
15726
  5%|▍ | 17/370 [00:03<01:18, 4.51it/s]
15727
  5%|▍ | 18/370 [00:04<01:13, 4.78it/s]
15728
  5%|▌ | 19/370 [00:04<01:17, 4.55it/s]
15729
  5%|▌ | 20/370 [00:04<01:19, 4.40it/s]
15730
  6%|▌ | 21/370 [00:04<01:16, 4.53it/s]
15731
  6%|▌ | 22/370 [00:05<01:18, 4.45it/s]
15732
  6%|▌ | 23/370 [00:05<01:15, 4.60it/s]
15733
  6%|▋ | 24/370 [00:05<01:16, 4.52it/s]
15734
  7%|▋ | 25/370 [00:05<01:18, 4.42it/s]
15735
  7%|▋ | 26/370 [00:05<01:13, 4.66it/s]
15736
  7%|▋ | 27/370 [00:06<01:13, 4.67it/s]
15737
  8%|▊ | 28/370 [00:06<01:14, 4.59it/s]
15738
  8%|▊ | 29/370 [00:06<01:18, 4.35it/s]
15739
  8%|▊ | 30/370 [00:06<01:21, 4.20it/s]
15740
  8%|▊ | 31/370 [00:07<01:20, 4.20it/s]
15741
  9%|▊ | 32/370 [00:07<01:18, 4.33it/s]
15742
  9%|▉ | 33/370 [00:07<01:16, 4.39it/s]
15743
  9%|▉ | 34/370 [00:07<01:18, 4.27it/s]
15744
  9%|▉ | 35/370 [00:08<01:23, 4.03it/s]
15745
  10%|▉ | 36/370 [00:08<01:21, 4.10it/s]
15746
  10%|█ | 37/370 [00:08<01:19, 4.20it/s]
15747
  10%|█ | 38/370 [00:08<01:18, 4.22it/s]
15748
  11%|█ | 39/370 [00:09<01:20, 4.10it/s]
15749
  11%|█ | 40/370 [00:09<01:21, 4.03it/s]
15750
  11%|█ | 41/370 [00:09<01:18, 4.20it/s]
15751
  11%|█▏ | 42/370 [00:09<01:20, 4.06it/s]
15752
  12%|█▏ | 43/370 [00:10<01:21, 4.03it/s]
15753
  12%|█▏ | 44/370 [00:10<01:17, 4.19it/s]
15754
  12%|█▏ | 45/370 [00:10<01:24, 3.84it/s]
15755
  12%|█▏ | 46/370 [00:10<01:21, 4.00it/s]
15756
  13%|█▎ | 47/370 [00:11<01:27, 3.68it/s]
15757
  13%|█▎ | 48/370 [00:11<01:29, 3.60it/s]
15758
  13%|█▎ | 49/370 [00:11<01:23, 3.83it/s]
15759
  14%|█▎ | 50/370 [00:11<01:16, 4.20it/s]
15760
  14%|█▍ | 51/370 [00:12<01:23, 3.81it/s]
15761
  14%|█▍ | 52/370 [00:12<01:16, 4.16it/s]
15762
  14%|█▍ | 53/370 [00:12<01:20, 3.93it/s]
15763
  15%|█▍ | 54/370 [00:12<01:16, 4.15it/s]
15764
  15%|█▍ | 55/370 [00:13<01:24, 3.73it/s]
15765
  15%|█▌ | 56/370 [00:13<01:22, 3.82it/s]
15766
  15%|█▌ | 57/370 [00:13<01:14, 4.17it/s]
15767
  16%|█▌ | 58/370 [00:13<01:20, 3.86it/s]
15768
  16%|█▌ | 59/370 [00:14<01:17, 3.99it/s]
15769
  16%|█▌ | 60/370 [00:14<01:12, 4.27it/s]
15770
  16%|█▋ | 61/370 [00:14<01:13, 4.18it/s]
15771
  17%|█▋ | 62/370 [00:14<01:14, 4.14it/s]
15772
  17%|█▋ | 63/370 [00:14<01:09, 4.42it/s]
15773
  17%|█▋ | 64/370 [00:15<01:07, 4.54it/s]
15774
  18%|█▊ | 65/370 [00:15<01:09, 4.36it/s]
15775
  18%|█▊ | 66/370 [00:15<01:08, 4.45it/s]
15776
  18%|█▊ | 67/370 [00:15<01:11, 4.25it/s]
15777
  18%|█▊ | 68/370 [00:16<01:16, 3.97it/s]
15778
  19%|█▊ | 69/370 [00:16<01:11, 4.23it/s]
15779
  19%|█▉ | 70/370 [00:16<01:09, 4.35it/s]
15780
  19%|█▉ | 71/370 [00:16<01:12, 4.10it/s]
15781
  19%|█▉ | 72/370 [00:17<01:06, 4.46it/s]
15782
  20%|█▉ | 73/370 [00:17<01:06, 4.47it/s]
15783
  20%|██ | 74/370 [00:17<01:12, 4.10it/s]
15784
  20%|██ | 75/370 [00:17<01:20, 3.67it/s]
15785
  21%|██ | 76/370 [00:18<01:12, 4.04it/s]
15786
  21%|██ | 77/370 [00:18<01:13, 3.98it/s]
15787
  21%|██ | 78/370 [00:18<01:11, 4.06it/s]
15788
  21%|██▏ | 79/370 [00:18<01:14, 3.93it/s]
15789
  22%|██▏ | 80/370 [00:19<01:10, 4.10it/s]
15790
  22%|██▏ | 81/370 [00:19<01:06, 4.33it/s]
15791
  22%|██▏ | 82/370 [00:19<01:05, 4.41it/s]
15792
  22%|██▏ | 83/370 [00:19<01:05, 4.37it/s]
15793
  23%|██▎ | 84/370 [00:19<01:06, 4.32it/s]
15794
  23%|██▎ | 85/370 [00:20<01:03, 4.47it/s]
15795
  23%|██▎ | 86/370 [00:20<01:02, 4.51it/s]
15796
  24%|██▎ | 87/370 [00:20<01:06, 4.25it/s]
15797
  24%|██▍ | 88/370 [00:20<01:12, 3.88it/s]
15798
  24%|██▍ | 89/370 [00:21<01:19, 3.53it/s]
15799
  24%|██▍ | 90/370 [00:21<01:12, 3.84it/s]
15800
  25%|██▍ | 91/370 [00:21<01:08, 4.07it/s]
15801
  25%|██▍ | 92/370 [00:22<01:14, 3.71it/s]
15802
  25%|██▌ | 93/370 [00:22<01:18, 3.51it/s]
15803
  25%|██▌ | 94/370 [00:22<01:13, 3.73it/s]
15804
  26%|██▌ | 95/370 [00:22<01:19, 3.45it/s]
15805
  26%|██▌ | 96/370 [00:23<01:16, 3.58it/s]
15806
  26%|██▌ | 97/370 [00:23<01:11, 3.82it/s]
15807
  26%|██▋ | 98/370 [00:23<01:08, 3.98it/s]
15808
  27%|██▋ | 99/370 [00:23<01:06, 4.10it/s]
15809
  27%|██▋ | 100/370 [00:24<01:02, 4.33it/s]
15810
  27%|██▋ | 101/370 [00:24<00:59, 4.49it/s]
15811
  28%|██▊ | 102/370 [00:24<00:57, 4.65it/s]
15812
  28%|██▊ | 103/370 [00:24<00:56, 4.70it/s]
15813
  28%|██▊ | 104/370 [00:25<01:03, 4.21it/s]
15814
  28%|██▊ | 105/370 [00:25<01:05, 4.06it/s]
15815
  29%|██▊ | 106/370 [00:25<01:04, 4.07it/s]
15816
  29%|██▉ | 107/370 [00:25<01:02, 4.19it/s]
15817
  29%|██▉ | 108/370 [00:26<01:05, 4.00it/s]
15818
  29%|██▉ | 109/370 [00:26<01:04, 4.02it/s]
15819
  30%|██▉ | 110/370 [00:26<01:00, 4.31it/s]
15820
  30%|███ | 111/370 [00:26<01:00, 4.27it/s]
15821
  30%|███ | 112/370 [00:26<01:01, 4.21it/s]
15822
  31%|███ | 113/370 [00:27<00:59, 4.33it/s]
15823
  31%|███ | 114/370 [00:27<00:57, 4.46it/s]
15824
  31%|███ | 115/370 [00:27<00:59, 4.25it/s]
15825
  31%|███▏ | 116/370 [00:27<01:02, 4.04it/s]
15826
  32%|███▏ | 117/370 [00:28<01:06, 3.78it/s]
15827
  32%|███▏ | 118/370 [00:28<01:04, 3.90it/s]
15828
  32%|███▏ | 119/370 [00:28<01:02, 4.05it/s]
15829
  32%|███▏ | 120/370 [00:28<00:59, 4.18it/s]
15830
  33%|███▎ | 121/370 [00:29<00:58, 4.26it/s]
15831
  33%|███▎ | 122/370 [00:29<00:57, 4.32it/s]
15832
  33%|███▎ | 123/370 [00:29<00:58, 4.25it/s]
15833
  34%|███▎ | 124/370 [00:29<00:54, 4.53it/s]
15834
  34%|███▍ | 125/370 [00:30<00:55, 4.38it/s]
15835
  34%|███▍ | 126/370 [00:30<01:00, 4.04it/s]
15836
  34%|███▍ | 127/370 [00:30<01:00, 4.03it/s]
15837
  35%|███▍ | 128/370 [00:30<01:04, 3.78it/s]
15838
  35%|███▍ | 129/370 [00:31<01:05, 3.66it/s]
15839
  35%|███▌ | 130/370 [00:31<01:05, 3.65it/s]
15840
  35%|███▌ | 131/370 [00:31<01:12, 3.31it/s]
15841
  36%|███▌ | 132/370 [00:31<01:04, 3.71it/s]
15842
  36%|███▌ | 133/370 [00:32<01:02, 3.81it/s]
15843
  36%|███▌ | 134/370 [00:32<00:59, 3.94it/s]
15844
  36%|███▋ | 135/370 [00:32<01:00, 3.89it/s]
15845
  37%|███▋ | 136/370 [00:32<00:58, 3.98it/s]
15846
  37%|███▋ | 137/370 [00:33<00:57, 4.04it/s]
15847
  37%|███▋ | 138/370 [00:33<00:56, 4.10it/s]
15848
  38%|███▊ | 139/370 [00:33<00:59, 3.88it/s]
15849
  38%|███▊ | 140/370 [00:33<00:55, 4.15it/s]
15850
  38%|███▊ | 141/370 [00:34<01:00, 3.77it/s]
15851
  38%|███▊ | 142/370 [00:34<00:59, 3.80it/s]
15852
  39%|███▊ | 143/370 [00:34<00:59, 3.84it/s]
15853
  39%|███▉ | 144/370 [00:35<00:59, 3.83it/s]
15854
  39%|███▉ | 145/370 [00:35<01:00, 3.70it/s]
15855
  39%|███▉ | 146/370 [00:35<00:56, 3.93it/s]
15856
  40%|███▉ | 147/370 [00:35<00:55, 4.00it/s]
15857
  40%|████ | 148/370 [00:36<00:57, 3.89it/s]
15858
  40%|████ | 149/370 [00:36<00:56, 3.94it/s]
15859
  41%|████ | 150/370 [00:36<00:50, 4.33it/s]
15860
  41%|████ | 151/370 [00:36<00:48, 4.54it/s]
15861
  41%|████ | 152/370 [00:36<00:49, 4.43it/s]
15862
  41%|████▏ | 153/370 [00:37<00:51, 4.21it/s]
15863
  42%|████▏ | 154/370 [00:37<00:51, 4.17it/s]
15864
  42%|████▏ | 155/370 [00:37<00:50, 4.23it/s]
15865
  42%|████▏ | 156/370 [00:37<00:54, 3.91it/s]
15866
  42%|████▏ | 157/370 [00:38<00:54, 3.93it/s]
15867
  43%|████▎ | 158/370 [00:38<00:52, 4.02it/s]
15868
  43%|████▎ | 159/370 [00:38<00:49, 4.29it/s]
15869
  43%|████▎ | 160/370 [00:38<00:49, 4.22it/s]
15870
  44%|████▎ | 161/370 [00:39<00:49, 4.22it/s]
15871
  44%|████▍ | 162/370 [00:39<00:50, 4.16it/s]
15872
  44%|████▍ | 163/370 [00:39<00:47, 4.35it/s]
15873
  44%|████▍ | 164/370 [00:39<00:48, 4.28it/s]
15874
  45%|████▍ | 165/370 [00:40<00:49, 4.10it/s]
15875
  45%|████▍ | 166/370 [00:40<00:50, 4.07it/s]
15876
  45%|████▌ | 167/370 [00:40<00:49, 4.11it/s]
15877
  45%|████▌ | 168/370 [00:40<00:48, 4.16it/s]
15878
  46%|████▌ | 169/370 [00:41<00:47, 4.26it/s]
15879
  46%|████▌ | 170/370 [00:41<00:47, 4.22it/s]
15880
  46%|████▌ | 171/370 [00:41<00:48, 4.09it/s]
15881
  46%|████▋ | 172/370 [00:41<00:44, 4.44it/s]
15882
  47%|████▋ | 173/370 [00:41<00:40, 4.82it/s]
15883
  47%|████▋ | 174/370 [00:42<00:45, 4.28it/s]
15884
  47%|████▋ | 175/370 [00:42<00:43, 4.45it/s]
15885
  48%|████▊ | 176/370 [00:42<00:41, 4.62it/s]
15886
  48%|████▊ | 177/370 [00:42<00:44, 4.37it/s]
15887
  48%|████▊ | 178/370 [00:43<00:44, 4.33it/s]
15888
  48%|████▊ | 179/370 [00:43<00:44, 4.33it/s]
15889
  49%|████▊ | 180/370 [00:43<00:40, 4.74it/s]
15890
  49%|████▉ | 181/370 [00:43<00:43, 4.30it/s]
15891
  49%|████▉ | 182/370 [00:43<00:42, 4.41it/s]
15892
  49%|████▉ | 183/370 [00:44<00:47, 3.95it/s]
15893
  50%|████▉ | 184/370 [00:44<00:45, 4.05it/s]
15894
  50%|█████ | 185/370 [00:44<00:45, 4.03it/s]
15895
  50%|█████ | 186/370 [00:44<00:44, 4.10it/s]
15896
  51%|█████ | 187/370 [00:45<00:45, 4.03it/s]
15897
  51%|█████ | 188/370 [00:45<00:44, 4.06it/s]
15898
  51%|█████ | 189/370 [00:45<00:46, 3.93it/s]
15899
  51%|█████▏ | 190/370 [00:46<00:47, 3.79it/s]
15900
  52%|█████▏ | 191/370 [00:46<00:44, 4.03it/s]
15901
  52%|█████▏ | 192/370 [00:46<00:49, 3.60it/s]
15902
  52%|█████▏ | 193/370 [00:46<00:48, 3.66it/s]
15903
  52%|█████▏ | 194/370 [00:47<00:42, 4.12it/s]
15904
  53%|█████▎ | 195/370 [00:47<00:42, 4.10it/s]
15905
  53%|█████▎ | 196/370 [00:47<00:41, 4.21it/s]
15906
  53%|█████▎ | 197/370 [00:47<00:40, 4.28it/s]
15907
  54%|█████▎ | 198/370 [00:47<00:40, 4.26it/s]
15908
  54%|█████▍ | 199/370 [00:48<00:40, 4.19it/s]
15909
  54%|█████▍ | 200/370 [00:48<00:47, 3.62it/s]
15910
  54%|█████▍ | 201/370 [00:48<00:44, 3.82it/s]
15911
  55%|█████▍ | 202/370 [00:49<00:43, 3.82it/s]
15912
  55%|█████▍ | 203/370 [00:49<00:42, 3.97it/s]
15913
  55%|█████▌ | 204/370 [00:49<00:43, 3.82it/s]
15914
  55%|█████▌ | 205/370 [00:49<00:40, 4.09it/s]
15915
  56%|█████▌ | 206/370 [00:50<00:42, 3.87it/s]
15916
  56%|█████▌ | 207/370 [00:50<00:41, 3.91it/s]
15917
  56%|█████▌ | 208/370 [00:50<00:43, 3.75it/s]
15918
  56%|█████▋ | 209/370 [00:50<00:42, 3.79it/s]
15919
  57%|█████▋ | 210/370 [00:51<00:43, 3.65it/s]
15920
  57%|█████▋ | 211/370 [00:51<00:44, 3.57it/s]
15921
  57%|█████▋ | 212/370 [00:51<00:43, 3.66it/s]
15922
  58%|█████▊ | 213/370 [00:52<00:43, 3.60it/s]
15923
  58%|█████▊ | 214/370 [00:52<00:40, 3.83it/s]
15924
  58%|█████▊ | 215/370 [00:52<00:40, 3.85it/s]
15925
  58%|█████▊ | 216/370 [00:52<00:39, 3.94it/s]
15926
  59%|█████▊ | 217/370 [00:53<00:40, 3.73it/s]
15927
  59%|█████▉ | 218/370 [00:53<00:37, 4.02it/s]
15928
  59%|█████▉ | 219/370 [00:53<00:38, 3.97it/s]
15929
  59%|█████▉ | 220/370 [00:53<00:37, 4.03it/s]
15930
  60%|█████▉ | 221/370 [00:54<00:37, 3.93it/s]
15931
  60%|██████ | 222/370 [00:54<00:36, 4.08it/s]
15932
  60%|██████ | 223/370 [00:54<00:37, 3.91it/s]
15933
  61%|██████ | 224/370 [00:54<00:40, 3.64it/s]
15934
  61%|██████ | 225/370 [00:55<00:42, 3.40it/s]
15935
  61%|██████ | 226/370 [00:55<00:40, 3.55it/s]
15936
  61%|██████▏ | 227/370 [00:55<00:40, 3.52it/s]
15937
  62%|██████▏ | 228/370 [00:56<00:40, 3.49it/s]
15938
  62%|██████▏ | 229/370 [00:56<00:40, 3.49it/s]
15939
  62%|██████▏ | 230/370 [00:56<00:39, 3.51it/s]
15940
  62%|██████▏ | 231/370 [00:56<00:41, 3.38it/s]
15941
  63%|██████▎ | 232/370 [00:57<00:41, 3.32it/s]
15942
  63%|██████▎ | 233/370 [00:57<00:41, 3.31it/s]
15943
  63%|██████▎ | 234/370 [00:57<00:39, 3.46it/s]
15944
  64%|██████▎ | 235/370 [00:58<00:37, 3.61it/s]
15945
  64%|██████▍ | 236/370 [00:58<00:35, 3.79it/s]
15946
  64%|██████▍ | 237/370 [00:58<00:35, 3.75it/s]
15947
  64%|██████▍ | 238/370 [00:58<00:36, 3.62it/s]
15948
  65%|██████▍ | 239/370 [00:59<00:35, 3.69it/s]
15949
  65%|██████▍ | 240/370 [00:59<00:36, 3.52it/s]
15950
  65%|██████▌ | 241/370 [00:59<00:37, 3.46it/s]
15951
  65%|██████▌ | 242/370 [00:59<00:35, 3.56it/s]
15952
  66%|██████▌ | 243/370 [01:00<00:38, 3.28it/s]
15953
  66%|██████▌ | 244/370 [01:00<00:40, 3.08it/s]
15954
  66%|██████▌ | 245/370 [01:00<00:38, 3.21it/s]
15955
  66%|██████▋ | 246/370 [01:01<00:36, 3.36it/s]
15956
  67%|██████▋ | 247/370 [01:01<00:33, 3.67it/s]
15957
  67%|██████▋ | 248/370 [01:01<00:32, 3.80it/s]
15958
  67%|██████▋ | 249/370 [01:01<00:32, 3.75it/s]
15959
  68%|██████▊ | 250/370 [01:02<00:33, 3.62it/s]
15960
  68%|██████▊ | 251/370 [01:02<00:32, 3.68it/s]
15961
  68%|██████▊ | 252/370 [01:02<00:36, 3.26it/s]
15962
  68%|██████▊ | 253/370 [01:03<00:35, 3.31it/s]
15963
  69%|██████▊ | 254/370 [01:03<00:37, 3.13it/s]
15964
  69%|██████▉ | 255/370 [01:03<00:36, 3.19it/s]
15965
  69%|██████▉ | 256/370 [01:04<00:36, 3.14it/s]
15966
  69%|██████▉ | 257/370 [01:04<00:36, 3.09it/s]
15967
  70%|██████▉ | 258/370 [01:04<00:35, 3.13it/s]
15968
  70%|███████ | 259/370 [01:05<00:34, 3.19it/s]
15969
  70%|███████ | 260/370 [01:05<00:32, 3.39it/s]
15970
  71%|███████ | 261/370 [01:05<00:32, 3.32it/s]
15971
  71%|███████ | 262/370 [01:06<00:32, 3.31it/s]
15972
  71%|███████ | 263/370 [01:06<00:33, 3.20it/s]
15973
  71%|███████▏ | 264/370 [01:06<00:34, 3.09it/s]
15974
  72%|███████▏ | 265/370 [01:07<00:33, 3.17it/s]
15975
  72%|███████▏ | 266/370 [01:07<00:33, 3.10it/s]
15976
  72%|███████▏ | 267/370 [01:07<00:33, 3.04it/s]
15977
  72%|███████▏ | 268/370 [01:08<00:33, 3.05it/s]
15978
  73%|███████▎ | 269/370 [01:08<00:33, 2.97it/s]
15979
  73%|███████▎ | 270/370 [01:08<00:33, 3.00it/s]
15980
  73%|███████▎ | 271/370 [01:09<00:33, 2.94it/s]
15981
  74%|███████▎ | 272/370 [01:09<00:31, 3.15it/s]
15982
  74%|███████▍ | 273/370 [01:09<00:31, 3.12it/s]
15983
  74%|███████▍ | 274/370 [01:09<00:31, 3.03it/s]
15984
  74%|███████▍ | 275/370 [01:10<00:30, 3.16it/s]
15985
  75%|███████▍ | 276/370 [01:10<00:29, 3.19it/s]
15986
  75%|███████▍ | 277/370 [01:10<00:31, 2.97it/s]
15987
  75%|███████▌ | 278/370 [01:11<00:31, 2.94it/s]
15988
  75%|███████▌ | 279/370 [01:11<00:30, 2.97it/s]
15989
  76%|███████▌ | 280/370 [01:12<00:31, 2.90it/s]
15990
  76%|███████▌ | 281/370 [01:12<00:29, 2.98it/s]
15991
  76%|███████▌ | 282/370 [01:12<00:29, 2.98it/s]
15992
  76%|███████▋ | 283/370 [01:13<00:30, 2.81it/s]
15993
  77%|███████▋ | 284/370 [01:13<00:30, 2.83it/s]
15994
  77%|███████▋ | 285/370 [01:13<00:31, 2.68it/s]
15995
  77%|███████▋ | 286/370 [01:14<00:30, 2.74it/s]
15996
  78%|███████▊ | 287/370 [01:14<00:28, 2.87it/s]
15997
  78%|███████▊ | 288/370 [01:14<00:27, 2.94it/s]
15998
  78%|███████▊ | 289/370 [01:15<00:26, 3.03it/s]
15999
  78%|███████▊ | 290/370 [01:15<00:26, 2.99it/s]
16000
  79%|███████▊ | 291/370 [01:15<00:26, 3.00it/s]
16001
  79%|███████▉ | 292/370 [01:16<00:24, 3.12it/s]
16002
  79%|███████▉ | 293/370 [01:16<00:24, 3.08it/s]
16003
  79%|███████▉ | 294/370 [01:16<00:24, 3.16it/s]
16004
  80%|███████▉ | 295/370 [01:16<00:22, 3.32it/s]
16005
  80%|████████ | 296/370 [01:17<00:21, 3.39it/s]
16006
  80%|████████ | 297/370 [01:17<00:22, 3.23it/s]
16007
  81%|████████ | 298/370 [01:17<00:23, 3.11it/s]
16008
  81%|████████ | 299/370 [01:18<00:23, 3.03it/s]
16009
  81%|████████ | 300/370 [01:18<00:22, 3.16it/s]
16010
  81%|████████▏ | 301/370 [01:18<00:22, 3.09it/s]
16011
  82%|████████▏ | 302/370 [01:19<00:20, 3.26it/s]
16012
  82%|████████▏ | 303/370 [01:19<00:20, 3.30it/s]
16013
  82%|████████▏ | 304/370 [01:19<00:20, 3.24it/s]
16014
  82%|████████▏ | 305/370 [01:20<00:20, 3.14it/s]
16015
  83%|████████▎ | 306/370 [01:20<00:20, 3.20it/s]
16016
  83%|████████▎ | 307/370 [01:20<00:19, 3.28it/s]
16017
  83%|████████▎ | 308/370 [01:21<00:18, 3.36it/s]
16018
  84%|████████▎ | 309/370 [01:21<00:19, 3.18it/s]
16019
  84%|████████▍ | 310/370 [01:21<00:19, 3.05it/s]
16020
  84%|████████▍ | 311/370 [01:22<00:18, 3.11it/s]
16021
  84%|████████▍ | 312/370 [01:22<00:19, 3.04it/s]
16022
  85%|████████▍ | 313/370 [01:22<00:17, 3.17it/s]
16023
  85%|████████▍ | 314/370 [01:22<00:16, 3.36it/s]
16024
  85%|████████▌ | 315/370 [01:23<00:16, 3.43it/s]
16025
  85%|████████▌ | 316/370 [01:23<00:15, 3.40it/s]
16026
  86%|████████▌ | 317/370 [01:23<00:16, 3.28it/s]
16027
  86%|████████▌ | 318/370 [01:24<00:16, 3.10it/s]
16028
  86%|████████▌ | 319/370 [01:24<00:15, 3.24it/s]
16029
  86%|████████▋ | 320/370 [01:24<00:15, 3.32it/s]
16030
  87%|████████▋ | 321/370 [01:25<00:14, 3.37it/s]
16031
  87%|████████▋ | 322/370 [01:25<00:14, 3.40it/s]
16032
  87%|████████▋ | 323/370 [01:25<00:13, 3.46it/s]
16033
  88%|████████▊ | 324/370 [01:25<00:13, 3.33it/s]
16034
  88%|████████▊ | 325/370 [01:26<00:13, 3.39it/s]
16035
  88%|████████▊ | 326/370 [01:26<00:13, 3.37it/s]
16036
  88%|████████▊ | 327/370 [01:26<00:12, 3.31it/s]
16037
  89%|████████▊ | 328/370 [01:27<00:12, 3.42it/s]
16038
  89%|████████▉ | 329/370 [01:27<00:11, 3.54it/s]
16039
  89%|████████▉ | 330/370 [01:27<00:12, 3.25it/s]
16040
  89%|████████▉ | 331/370 [01:28<00:11, 3.27it/s]
16041
  90%|████████▉ | 332/370 [01:28<00:11, 3.38it/s]
16042
  90%|█████████ | 333/370 [01:28<00:11, 3.33it/s]
16043
  90%|█████████ | 334/370 [01:28<00:10, 3.36it/s]
16044
  91%|█████████ | 335/370 [01:29<00:09, 3.50it/s]
16045
  91%|█████████ | 336/370 [01:29<00:09, 3.57it/s]
16046
  91%|█████████ | 337/370 [01:29<00:09, 3.51it/s]
16047
  91%|█████████▏| 338/370 [01:30<00:09, 3.39it/s]
16048
  92%|█████████▏| 339/370 [01:30<00:08, 3.46it/s]
16049
  92%|█████████▏| 340/370 [01:30<00:08, 3.36it/s]
16050
  92%|█████████▏| 341/370 [01:30<00:08, 3.28it/s]
16051
  92%|█████████▏| 342/370 [01:31<00:08, 3.43it/s]
16052
  93%|█████████▎| 343/370 [01:31<00:07, 3.40it/s]
16053
  93%|█████████▎| 344/370 [01:31<00:08, 3.09it/s]
16054
  93%|█████████▎| 345/370 [01:32<00:08, 2.87it/s]
16055
  94%|█████████▎| 346/370 [01:32<00:08, 2.90it/s]
16056
  94%|█████████▍| 347/370 [01:33<00:07, 2.88it/s]
16057
  94%|█████████▍| 348/370 [01:33<00:07, 3.03it/s]
16058
  94%|█████████▍| 349/370 [01:33<00:06, 3.06it/s]
16059
  95%|█████████▍| 350/370 [01:33<00:06, 3.18it/s]
16060
  95%|█████████▍| 351/370 [01:34<00:06, 3.12it/s]
16061
  95%|█████████▌| 352/370 [01:34<00:05, 3.06it/s]
16062
  95%|█████████▌| 353/370 [01:34<00:05, 3.08it/s]
16063
  96%|█████████▌| 354/370 [01:35<00:05, 3.01it/s]
16064
  96%|█████████▌| 355/370 [01:35<00:04, 3.24it/s]
16065
  96%|█████████▌| 356/370 [01:35<00:04, 3.43it/s]
16066
  96%|█████████▋| 357/370 [01:36<00:03, 3.58it/s]
16067
  97%|█████████▋| 358/370 [01:36<00:03, 3.88it/s]
16068
  97%|█████████▋| 359/370 [01:36<00:02, 3.93it/s]
16069
  97%|█████████▋| 360/370 [01:36<00:02, 3.61it/s]
16070
  98%|█████████▊| 361/370 [01:37<00:02, 3.70it/s]
16071
  98%|█████████▊| 362/370 [01:37<00:02, 3.78it/s]
16072
  98%|█████████▊| 363/370 [01:37<00:01, 3.73it/s]
16073
  98%|█████████▊| 364/370 [01:37<00:01, 3.56it/s]
16074
  99%|█████████▊| 365/370 [01:38<00:01, 3.81it/s]
16075
  99%|█████████▉| 366/370 [01:38<00:01, 3.57it/s]
16076
  99%|█████████▉| 367/370 [01:38<00:00, 3.69it/s]
16077
  99%|█████████▉| 368/370 [01:38<00:00, 3.73it/s]
 
 
 
 
15707
 
15708
  Dropping the following result as it does not have all the necessary fields:
15709
  {'dataset': {'name': 'common_voice', 'type': 'common_voice', 'args': 'zh-TW'}}
15710
+ To https://huggingface.co/StevenLimcorn/wav2vec2-xls-r-300m-zh-TW
15711
+ 97bab54..a214e6b main -> main
15712
+
15713
+ 02/06/2022 21:54:37 - WARNING - huggingface_hub.repository - To https://huggingface.co/StevenLimcorn/wav2vec2-xls-r-300m-zh-TW
15714
+ 97bab54..a214e6b main -> main
15715
+
15716
+ ***** train metrics *****
15717
+ epoch = 100.0
15718
+ train_loss = 5.9196
15719
+ train_runtime = 10:09:47.17
15720
+ train_samples = 6380
15721
+ train_samples_per_second = 17.438
15722
+ train_steps_per_second = 0.544
15723
+ 02/06/2022 21:54:39 - INFO - __main__ - *** Evaluate ***
15724
+ The following columns in the evaluation set don't have a corresponding argument in `Wav2Vec2ForCTC.forward` and have been ignored: input_length.
15725
+ ***** Running Evaluation *****
15726
+ Num examples = 2958
15727
+ Batch size = 8
15728
+
15729
  0%| | 0/370 [00:00<?, ?it/s]
15730
  1%| | 2/370 [00:00<00:43, 8.45it/s]
15731
  1%| | 3/370 [00:00<01:04, 5.69it/s]
15732
  1%| | 4/370 [00:00<01:15, 4.82it/s]
15733
  1%|▏ | 5/370 [00:01<01:26, 4.23it/s]
15734
  2%|▏ | 6/370 [00:01<01:28, 4.13it/s]
15735
  2%|▏ | 7/370 [00:01<01:32, 3.93it/s]
15736
  2%|▏ | 8/370 [00:01<01:36, 3.76it/s]
15737
  2%|▏ | 9/370 [00:02<01:32, 3.92it/s]
15738
  3%|▎ | 10/370 [00:02<01:33, 3.84it/s]
15739
  3%|▎ | 11/370 [00:02<01:29, 4.00it/s]
15740
  3%|▎ | 12/370 [00:02<01:27, 4.09it/s]
15741
  4%|▎ | 13/370 [00:03<01:25, 4.16it/s]
15742
  4%|▍ | 14/370 [00:03<01:28, 4.02it/s]
15743
  4%|▍ | 15/370 [00:03<01:22, 4.33it/s]
15744
  4%|▍ | 16/370 [00:03<01:23, 4.22it/s]
15745
  5%|▍ | 17/370 [00:03<01:18, 4.51it/s]
15746
  5%|▍ | 18/370 [00:04<01:13, 4.78it/s]
15747
  5%|▌ | 19/370 [00:04<01:17, 4.55it/s]
15748
  5%|▌ | 20/370 [00:04<01:19, 4.40it/s]
15749
  6%|▌ | 21/370 [00:04<01:16, 4.53it/s]
15750
  6%|▌ | 22/370 [00:05<01:18, 4.45it/s]
15751
  6%|▌ | 23/370 [00:05<01:15, 4.60it/s]
15752
  6%|▋ | 24/370 [00:05<01:16, 4.52it/s]
15753
  7%|▋ | 25/370 [00:05<01:18, 4.42it/s]
15754
  7%|▋ | 26/370 [00:05<01:13, 4.66it/s]
15755
  7%|▋ | 27/370 [00:06<01:13, 4.67it/s]
15756
  8%|▊ | 28/370 [00:06<01:14, 4.59it/s]
15757
  8%|▊ | 29/370 [00:06<01:18, 4.35it/s]
15758
  8%|▊ | 30/370 [00:06<01:21, 4.20it/s]
15759
  8%|▊ | 31/370 [00:07<01:20, 4.20it/s]
15760
  9%|▊ | 32/370 [00:07<01:18, 4.33it/s]
15761
  9%|▉ | 33/370 [00:07<01:16, 4.39it/s]
15762
  9%|▉ | 34/370 [00:07<01:18, 4.27it/s]
15763
  9%|▉ | 35/370 [00:08<01:23, 4.03it/s]
15764
  10%|▉ | 36/370 [00:08<01:21, 4.10it/s]
15765
  10%|█ | 37/370 [00:08<01:19, 4.20it/s]
15766
  10%|█ | 38/370 [00:08<01:18, 4.22it/s]
15767
  11%|█ | 39/370 [00:09<01:20, 4.10it/s]
15768
  11%|█ | 40/370 [00:09<01:21, 4.03it/s]
15769
  11%|█ | 41/370 [00:09<01:18, 4.20it/s]
15770
  11%|█▏ | 42/370 [00:09<01:20, 4.06it/s]
15771
  12%|█▏ | 43/370 [00:10<01:21, 4.03it/s]
15772
  12%|█▏ | 44/370 [00:10<01:17, 4.19it/s]
15773
  12%|█▏ | 45/370 [00:10<01:24, 3.84it/s]
15774
  12%|█▏ | 46/370 [00:10<01:21, 4.00it/s]
15775
  13%|█▎ | 47/370 [00:11<01:27, 3.68it/s]
15776
  13%|█▎ | 48/370 [00:11<01:29, 3.60it/s]
15777
  13%|█▎ | 49/370 [00:11<01:23, 3.83it/s]
15778
  14%|█▎ | 50/370 [00:11<01:16, 4.20it/s]
15779
  14%|█▍ | 51/370 [00:12<01:23, 3.81it/s]
15780
  14%|█▍ | 52/370 [00:12<01:16, 4.16it/s]
15781
  14%|█▍ | 53/370 [00:12<01:20, 3.93it/s]
15782
  15%|█▍ | 54/370 [00:12<01:16, 4.15it/s]
15783
  15%|█▍ | 55/370 [00:13<01:24, 3.73it/s]
15784
  15%|█▌ | 56/370 [00:13<01:22, 3.82it/s]
15785
  15%|█▌ | 57/370 [00:13<01:14, 4.17it/s]
15786
  16%|█▌ | 58/370 [00:13<01:20, 3.86it/s]
15787
  16%|█▌ | 59/370 [00:14<01:17, 3.99it/s]
15788
  16%|█▌ | 60/370 [00:14<01:12, 4.27it/s]
15789
  16%|█▋ | 61/370 [00:14<01:13, 4.18it/s]
15790
  17%|█▋ | 62/370 [00:14<01:14, 4.14it/s]
15791
  17%|█▋ | 63/370 [00:14<01:09, 4.42it/s]
15792
  17%|█▋ | 64/370 [00:15<01:07, 4.54it/s]
15793
  18%|█▊ | 65/370 [00:15<01:09, 4.36it/s]
15794
  18%|█▊ | 66/370 [00:15<01:08, 4.45it/s]
15795
  18%|█▊ | 67/370 [00:15<01:11, 4.25it/s]
15796
  18%|█▊ | 68/370 [00:16<01:16, 3.97it/s]
15797
  19%|█▊ | 69/370 [00:16<01:11, 4.23it/s]
15798
  19%|█▉ | 70/370 [00:16<01:09, 4.35it/s]
15799
  19%|█▉ | 71/370 [00:16<01:12, 4.10it/s]
15800
  19%|█▉ | 72/370 [00:17<01:06, 4.46it/s]
15801
  20%|█▉ | 73/370 [00:17<01:06, 4.47it/s]
15802
  20%|██ | 74/370 [00:17<01:12, 4.10it/s]
15803
  20%|██ | 75/370 [00:17<01:20, 3.67it/s]
15804
  21%|██ | 76/370 [00:18<01:12, 4.04it/s]
15805
  21%|██ | 77/370 [00:18<01:13, 3.98it/s]
15806
  21%|██ | 78/370 [00:18<01:11, 4.06it/s]
15807
  21%|██▏ | 79/370 [00:18<01:14, 3.93it/s]
15808
  22%|██▏ | 80/370 [00:19<01:10, 4.10it/s]
15809
  22%|██▏ | 81/370 [00:19<01:06, 4.33it/s]
15810
  22%|██▏ | 82/370 [00:19<01:05, 4.41it/s]
15811
  22%|██▏ | 83/370 [00:19<01:05, 4.37it/s]
15812
  23%|██▎ | 84/370 [00:19<01:06, 4.32it/s]
15813
  23%|██▎ | 85/370 [00:20<01:03, 4.47it/s]
15814
  23%|██▎ | 86/370 [00:20<01:02, 4.51it/s]
15815
  24%|██▎ | 87/370 [00:20<01:06, 4.25it/s]
15816
  24%|██▍ | 88/370 [00:20<01:12, 3.88it/s]
15817
  24%|██▍ | 89/370 [00:21<01:19, 3.53it/s]
15818
  24%|██▍ | 90/370 [00:21<01:12, 3.84it/s]
15819
  25%|██▍ | 91/370 [00:21<01:08, 4.07it/s]
15820
  25%|██▍ | 92/370 [00:22<01:14, 3.71it/s]
15821
  25%|██▌ | 93/370 [00:22<01:18, 3.51it/s]
15822
  25%|██▌ | 94/370 [00:22<01:13, 3.73it/s]
15823
  26%|██▌ | 95/370 [00:22<01:19, 3.45it/s]
15824
  26%|██▌ | 96/370 [00:23<01:16, 3.58it/s]
15825
  26%|██▌ | 97/370 [00:23<01:11, 3.82it/s]
15826
  26%|██▋ | 98/370 [00:23<01:08, 3.98it/s]
15827
  27%|██▋ | 99/370 [00:23<01:06, 4.10it/s]
15828
  27%|██▋ | 100/370 [00:24<01:02, 4.33it/s]
15829
  27%|██▋ | 101/370 [00:24<00:59, 4.49it/s]
15830
  28%|██▊ | 102/370 [00:24<00:57, 4.65it/s]
15831
  28%|██▊ | 103/370 [00:24<00:56, 4.70it/s]
15832
  28%|██▊ | 104/370 [00:25<01:03, 4.21it/s]
15833
  28%|██▊ | 105/370 [00:25<01:05, 4.06it/s]
15834
  29%|██▊ | 106/370 [00:25<01:04, 4.07it/s]
15835
  29%|██▉ | 107/370 [00:25<01:02, 4.19it/s]
15836
  29%|██▉ | 108/370 [00:26<01:05, 4.00it/s]
15837
  29%|██▉ | 109/370 [00:26<01:04, 4.02it/s]
15838
  30%|██▉ | 110/370 [00:26<01:00, 4.31it/s]
15839
  30%|███ | 111/370 [00:26<01:00, 4.27it/s]
15840
  30%|███ | 112/370 [00:26<01:01, 4.21it/s]
15841
  31%|███ | 113/370 [00:27<00:59, 4.33it/s]
15842
  31%|███ | 114/370 [00:27<00:57, 4.46it/s]
15843
  31%|███ | 115/370 [00:27<00:59, 4.25it/s]
15844
  31%|███▏ | 116/370 [00:27<01:02, 4.04it/s]
15845
  32%|███▏ | 117/370 [00:28<01:06, 3.78it/s]
15846
  32%|███▏ | 118/370 [00:28<01:04, 3.90it/s]
15847
  32%|███▏ | 119/370 [00:28<01:02, 4.05it/s]
15848
  32%|███▏ | 120/370 [00:28<00:59, 4.18it/s]
15849
  33%|███▎ | 121/370 [00:29<00:58, 4.26it/s]
15850
  33%|███▎ | 122/370 [00:29<00:57, 4.32it/s]
15851
  33%|███▎ | 123/370 [00:29<00:58, 4.25it/s]
15852
  34%|███▎ | 124/370 [00:29<00:54, 4.53it/s]
15853
  34%|███▍ | 125/370 [00:30<00:55, 4.38it/s]
15854
  34%|███▍ | 126/370 [00:30<01:00, 4.04it/s]
15855
  34%|███▍ | 127/370 [00:30<01:00, 4.03it/s]
15856
  35%|███▍ | 128/370 [00:30<01:04, 3.78it/s]
15857
  35%|███▍ | 129/370 [00:31<01:05, 3.66it/s]
15858
  35%|███▌ | 130/370 [00:31<01:05, 3.65it/s]
15859
  35%|███▌ | 131/370 [00:31<01:12, 3.31it/s]
15860
  36%|███▌ | 132/370 [00:31<01:04, 3.71it/s]
15861
  36%|███▌ | 133/370 [00:32<01:02, 3.81it/s]
15862
  36%|███▌ | 134/370 [00:32<00:59, 3.94it/s]
15863
  36%|███▋ | 135/370 [00:32<01:00, 3.89it/s]
15864
  37%|███▋ | 136/370 [00:32<00:58, 3.98it/s]
15865
  37%|███▋ | 137/370 [00:33<00:57, 4.04it/s]
15866
  37%|███▋ | 138/370 [00:33<00:56, 4.10it/s]
15867
  38%|███▊ | 139/370 [00:33<00:59, 3.88it/s]
15868
  38%|███▊ | 140/370 [00:33<00:55, 4.15it/s]
15869
  38%|███▊ | 141/370 [00:34<01:00, 3.77it/s]
15870
  38%|███▊ | 142/370 [00:34<00:59, 3.80it/s]
15871
  39%|███▊ | 143/370 [00:34<00:59, 3.84it/s]
15872
  39%|███▉ | 144/370 [00:35<00:59, 3.83it/s]
15873
  39%|███▉ | 145/370 [00:35<01:00, 3.70it/s]
15874
  39%|███▉ | 146/370 [00:35<00:56, 3.93it/s]
15875
  40%|███▉ | 147/370 [00:35<00:55, 4.00it/s]
15876
  40%|████ | 148/370 [00:36<00:57, 3.89it/s]
15877
  40%|████ | 149/370 [00:36<00:56, 3.94it/s]
15878
  41%|████ | 150/370 [00:36<00:50, 4.33it/s]
15879
  41%|████ | 151/370 [00:36<00:48, 4.54it/s]
15880
  41%|████ | 152/370 [00:36<00:49, 4.43it/s]
15881
  41%|████▏ | 153/370 [00:37<00:51, 4.21it/s]
15882
  42%|████▏ | 154/370 [00:37<00:51, 4.17it/s]
15883
  42%|████▏ | 155/370 [00:37<00:50, 4.23it/s]
15884
  42%|████▏ | 156/370 [00:37<00:54, 3.91it/s]
15885
  42%|████▏ | 157/370 [00:38<00:54, 3.93it/s]
15886
  43%|████▎ | 158/370 [00:38<00:52, 4.02it/s]
15887
  43%|████▎ | 159/370 [00:38<00:49, 4.29it/s]
15888
  43%|████▎ | 160/370 [00:38<00:49, 4.22it/s]
15889
  44%|████▎ | 161/370 [00:39<00:49, 4.22it/s]
15890
  44%|████▍ | 162/370 [00:39<00:50, 4.16it/s]
15891
  44%|████▍ | 163/370 [00:39<00:47, 4.35it/s]
15892
  44%|████▍ | 164/370 [00:39<00:48, 4.28it/s]
15893
  45%|████▍ | 165/370 [00:40<00:49, 4.10it/s]
15894
  45%|████▍ | 166/370 [00:40<00:50, 4.07it/s]
15895
  45%|████▌ | 167/370 [00:40<00:49, 4.11it/s]
15896
  45%|████▌ | 168/370 [00:40<00:48, 4.16it/s]
15897
  46%|████▌ | 169/370 [00:41<00:47, 4.26it/s]
15898
  46%|████▌ | 170/370 [00:41<00:47, 4.22it/s]
15899
  46%|████▌ | 171/370 [00:41<00:48, 4.09it/s]
15900
  46%|████▋ | 172/370 [00:41<00:44, 4.44it/s]
15901
  47%|████▋ | 173/370 [00:41<00:40, 4.82it/s]
15902
  47%|████▋ | 174/370 [00:42<00:45, 4.28it/s]
15903
  47%|████▋ | 175/370 [00:42<00:43, 4.45it/s]
15904
  48%|████▊ | 176/370 [00:42<00:41, 4.62it/s]
15905
  48%|████▊ | 177/370 [00:42<00:44, 4.37it/s]
15906
  48%|████▊ | 178/370 [00:43<00:44, 4.33it/s]
15907
  48%|████▊ | 179/370 [00:43<00:44, 4.33it/s]
15908
  49%|████▊ | 180/370 [00:43<00:40, 4.74it/s]
15909
  49%|████▉ | 181/370 [00:43<00:43, 4.30it/s]
15910
  49%|████▉ | 182/370 [00:43<00:42, 4.41it/s]
15911
  49%|████▉ | 183/370 [00:44<00:47, 3.95it/s]
15912
  50%|████▉ | 184/370 [00:44<00:45, 4.05it/s]
15913
  50%|█████ | 185/370 [00:44<00:45, 4.03it/s]
15914
  50%|█████ | 186/370 [00:44<00:44, 4.10it/s]
15915
  51%|█████ | 187/370 [00:45<00:45, 4.03it/s]
15916
  51%|█████ | 188/370 [00:45<00:44, 4.06it/s]
15917
  51%|█████ | 189/370 [00:45<00:46, 3.93it/s]
15918
  51%|█████▏ | 190/370 [00:46<00:47, 3.79it/s]
15919
  52%|█████▏ | 191/370 [00:46<00:44, 4.03it/s]
15920
  52%|█████▏ | 192/370 [00:46<00:49, 3.60it/s]
15921
  52%|█████▏ | 193/370 [00:46<00:48, 3.66it/s]
15922
  52%|█████▏ | 194/370 [00:47<00:42, 4.12it/s]
15923
  53%|█████▎ | 195/370 [00:47<00:42, 4.10it/s]
15924
  53%|█████▎ | 196/370 [00:47<00:41, 4.21it/s]
15925
  53%|█████▎ | 197/370 [00:47<00:40, 4.28it/s]
15926
  54%|█████▎ | 198/370 [00:47<00:40, 4.26it/s]
15927
  54%|█████▍ | 199/370 [00:48<00:40, 4.19it/s]
15928
  54%|█████▍ | 200/370 [00:48<00:47, 3.62it/s]
15929
  54%|█████▍ | 201/370 [00:48<00:44, 3.82it/s]
15930
  55%|█████▍ | 202/370 [00:49<00:43, 3.82it/s]
15931
  55%|█████▍ | 203/370 [00:49<00:42, 3.97it/s]
15932
  55%|█████▌ | 204/370 [00:49<00:43, 3.82it/s]
15933
  55%|█████▌ | 205/370 [00:49<00:40, 4.09it/s]
15934
  56%|█████▌ | 206/370 [00:50<00:42, 3.87it/s]
15935
  56%|█████▌ | 207/370 [00:50<00:41, 3.91it/s]
15936
  56%|█████▌ | 208/370 [00:50<00:43, 3.75it/s]
15937
  56%|█████▋ | 209/370 [00:50<00:42, 3.79it/s]
15938
  57%|█████▋ | 210/370 [00:51<00:43, 3.65it/s]
15939
  57%|█████▋ | 211/370 [00:51<00:44, 3.57it/s]
15940
  57%|█████▋ | 212/370 [00:51<00:43, 3.66it/s]
15941
  58%|█████▊ | 213/370 [00:52<00:43, 3.60it/s]
15942
  58%|█████▊ | 214/370 [00:52<00:40, 3.83it/s]
15943
  58%|█████▊ | 215/370 [00:52<00:40, 3.85it/s]
15944
  58%|█████▊ | 216/370 [00:52<00:39, 3.94it/s]
15945
  59%|█████▊ | 217/370 [00:53<00:40, 3.73it/s]
15946
  59%|█████▉ | 218/370 [00:53<00:37, 4.02it/s]
15947
  59%|█████▉ | 219/370 [00:53<00:38, 3.97it/s]
15948
  59%|█████▉ | 220/370 [00:53<00:37, 4.03it/s]
15949
  60%|█████▉ | 221/370 [00:54<00:37, 3.93it/s]
15950
  60%|██████ | 222/370 [00:54<00:36, 4.08it/s]
15951
  60%|██████ | 223/370 [00:54<00:37, 3.91it/s]
15952
  61%|██████ | 224/370 [00:54<00:40, 3.64it/s]
15953
  61%|██████ | 225/370 [00:55<00:42, 3.40it/s]
15954
  61%|██████ | 226/370 [00:55<00:40, 3.55it/s]
15955
  61%|██████▏ | 227/370 [00:55<00:40, 3.52it/s]
15956
  62%|██████▏ | 228/370 [00:56<00:40, 3.49it/s]
15957
  62%|██████▏ | 229/370 [00:56<00:40, 3.49it/s]
15958
  62%|██████▏ | 230/370 [00:56<00:39, 3.51it/s]
15959
  62%|██████▏ | 231/370 [00:56<00:41, 3.38it/s]
15960
  63%|██████▎ | 232/370 [00:57<00:41, 3.32it/s]
15961
  63%|██████▎ | 233/370 [00:57<00:41, 3.31it/s]
15962
  63%|██████▎ | 234/370 [00:57<00:39, 3.46it/s]
15963
  64%|██████▎ | 235/370 [00:58<00:37, 3.61it/s]
15964
  64%|██████▍ | 236/370 [00:58<00:35, 3.79it/s]
15965
  64%|██████▍ | 237/370 [00:58<00:35, 3.75it/s]
15966
  64%|██████▍ | 238/370 [00:58<00:36, 3.62it/s]
15967
  65%|██████▍ | 239/370 [00:59<00:35, 3.69it/s]
15968
  65%|██████▍ | 240/370 [00:59<00:36, 3.52it/s]
15969
  65%|██████▌ | 241/370 [00:59<00:37, 3.46it/s]
15970
  65%|██████▌ | 242/370 [00:59<00:35, 3.56it/s]
15971
  66%|██████▌ | 243/370 [01:00<00:38, 3.28it/s]
15972
  66%|██████▌ | 244/370 [01:00<00:40, 3.08it/s]
15973
  66%|██████▌ | 245/370 [01:00<00:38, 3.21it/s]
15974
  66%|██████▋ | 246/370 [01:01<00:36, 3.36it/s]
15975
  67%|██████▋ | 247/370 [01:01<00:33, 3.67it/s]
15976
  67%|██████▋ | 248/370 [01:01<00:32, 3.80it/s]
15977
  67%|██████▋ | 249/370 [01:01<00:32, 3.75it/s]
15978
  68%|██████▊ | 250/370 [01:02<00:33, 3.62it/s]
15979
  68%|██████▊ | 251/370 [01:02<00:32, 3.68it/s]
15980
  68%|██████▊ | 252/370 [01:02<00:36, 3.26it/s]
15981
  68%|██████▊ | 253/370 [01:03<00:35, 3.31it/s]
15982
  69%|██████▊ | 254/370 [01:03<00:37, 3.13it/s]
15983
  69%|██████▉ | 255/370 [01:03<00:36, 3.19it/s]
15984
  69%|██████▉ | 256/370 [01:04<00:36, 3.14it/s]
15985
  69%|██████▉ | 257/370 [01:04<00:36, 3.09it/s]
15986
  70%|██████▉ | 258/370 [01:04<00:35, 3.13it/s]
15987
  70%|███████ | 259/370 [01:05<00:34, 3.19it/s]
15988
  70%|███████ | 260/370 [01:05<00:32, 3.39it/s]
15989
  71%|███████ | 261/370 [01:05<00:32, 3.32it/s]
15990
  71%|███████ | 262/370 [01:06<00:32, 3.31it/s]
15991
  71%|███████ | 263/370 [01:06<00:33, 3.20it/s]
15992
  71%|███████▏ | 264/370 [01:06<00:34, 3.09it/s]
15993
  72%|███████▏ | 265/370 [01:07<00:33, 3.17it/s]
15994
  72%|███████▏ | 266/370 [01:07<00:33, 3.10it/s]
15995
  72%|███████▏ | 267/370 [01:07<00:33, 3.04it/s]
15996
  72%|███████▏ | 268/370 [01:08<00:33, 3.05it/s]
15997
  73%|███████▎ | 269/370 [01:08<00:33, 2.97it/s]
15998
  73%|███████▎ | 270/370 [01:08<00:33, 3.00it/s]
15999
  73%|███████▎ | 271/370 [01:09<00:33, 2.94it/s]
16000
  74%|███████▎ | 272/370 [01:09<00:31, 3.15it/s]
16001
  74%|███████▍ | 273/370 [01:09<00:31, 3.12it/s]
16002
  74%|███████▍ | 274/370 [01:09<00:31, 3.03it/s]
16003
  74%|███████▍ | 275/370 [01:10<00:30, 3.16it/s]
16004
  75%|███████▍ | 276/370 [01:10<00:29, 3.19it/s]
16005
  75%|███████▍ | 277/370 [01:10<00:31, 2.97it/s]
16006
  75%|███████▌ | 278/370 [01:11<00:31, 2.94it/s]
16007
  75%|███████▌ | 279/370 [01:11<00:30, 2.97it/s]
16008
  76%|███████▌ | 280/370 [01:12<00:31, 2.90it/s]
16009
  76%|███████▌ | 281/370 [01:12<00:29, 2.98it/s]
16010
  76%|███████▌ | 282/370 [01:12<00:29, 2.98it/s]
16011
  76%|███████▋ | 283/370 [01:13<00:30, 2.81it/s]
16012
  77%|███████▋ | 284/370 [01:13<00:30, 2.83it/s]
16013
  77%|███████▋ | 285/370 [01:13<00:31, 2.68it/s]
16014
  77%|███████▋ | 286/370 [01:14<00:30, 2.74it/s]
16015
  78%|███████▊ | 287/370 [01:14<00:28, 2.87it/s]
16016
  78%|███████▊ | 288/370 [01:14<00:27, 2.94it/s]
16017
  78%|███████▊ | 289/370 [01:15<00:26, 3.03it/s]
16018
  78%|███████▊ | 290/370 [01:15<00:26, 2.99it/s]
16019
  79%|███████▊ | 291/370 [01:15<00:26, 3.00it/s]
16020
  79%|███████▉ | 292/370 [01:16<00:24, 3.12it/s]
16021
  79%|███████▉ | 293/370 [01:16<00:24, 3.08it/s]
16022
  79%|███████▉ | 294/370 [01:16<00:24, 3.16it/s]
16023
  80%|███████▉ | 295/370 [01:16<00:22, 3.32it/s]
16024
  80%|████████ | 296/370 [01:17<00:21, 3.39it/s]
16025
  80%|████████ | 297/370 [01:17<00:22, 3.23it/s]
16026
  81%|████████ | 298/370 [01:17<00:23, 3.11it/s]
16027
  81%|████████ | 299/370 [01:18<00:23, 3.03it/s]
16028
  81%|████████ | 300/370 [01:18<00:22, 3.16it/s]
16029
  81%|████████▏ | 301/370 [01:18<00:22, 3.09it/s]
16030
  82%|████████▏ | 302/370 [01:19<00:20, 3.26it/s]
16031
  82%|████████▏ | 303/370 [01:19<00:20, 3.30it/s]
16032
  82%|████████▏ | 304/370 [01:19<00:20, 3.24it/s]
16033
  82%|████████▏ | 305/370 [01:20<00:20, 3.14it/s]
16034
  83%|████████▎ | 306/370 [01:20<00:20, 3.20it/s]
16035
  83%|████████▎ | 307/370 [01:20<00:19, 3.28it/s]
16036
  83%|████████▎ | 308/370 [01:21<00:18, 3.36it/s]
16037
  84%|████████▎ | 309/370 [01:21<00:19, 3.18it/s]
16038
  84%|████████▍ | 310/370 [01:21<00:19, 3.05it/s]
16039
  84%|████████▍ | 311/370 [01:22<00:18, 3.11it/s]
16040
  84%|████████▍ | 312/370 [01:22<00:19, 3.04it/s]
16041
  85%|████████▍ | 313/370 [01:22<00:17, 3.17it/s]
16042
  85%|████████▍ | 314/370 [01:22<00:16, 3.36it/s]
16043
  85%|████████▌ | 315/370 [01:23<00:16, 3.43it/s]
16044
  85%|████████▌ | 316/370 [01:23<00:15, 3.40it/s]
16045
  86%|████████▌ | 317/370 [01:23<00:16, 3.28it/s]
16046
  86%|████████▌ | 318/370 [01:24<00:16, 3.10it/s]
16047
  86%|████████▌ | 319/370 [01:24<00:15, 3.24it/s]
16048
  86%|████████▋ | 320/370 [01:24<00:15, 3.32it/s]
16049
  87%|████████▋ | 321/370 [01:25<00:14, 3.37it/s]
16050
  87%|████████▋ | 322/370 [01:25<00:14, 3.40it/s]
16051
  87%|████████▋ | 323/370 [01:25<00:13, 3.46it/s]
16052
  88%|████████▊ | 324/370 [01:25<00:13, 3.33it/s]
16053
  88%|████████▊ | 325/370 [01:26<00:13, 3.39it/s]
16054
  88%|████████▊ | 326/370 [01:26<00:13, 3.37it/s]
16055
  88%|████████▊ | 327/370 [01:26<00:12, 3.31it/s]
16056
  89%|████████▊ | 328/370 [01:27<00:12, 3.42it/s]
16057
  89%|████████▉ | 329/370 [01:27<00:11, 3.54it/s]
16058
  89%|████████▉ | 330/370 [01:27<00:12, 3.25it/s]
16059
  89%|████████▉ | 331/370 [01:28<00:11, 3.27it/s]
16060
  90%|████████▉ | 332/370 [01:28<00:11, 3.38it/s]
16061
  90%|█████████ | 333/370 [01:28<00:11, 3.33it/s]
16062
  90%|█████████ | 334/370 [01:28<00:10, 3.36it/s]
16063
  91%|█████████ | 335/370 [01:29<00:09, 3.50it/s]
16064
  91%|█████████ | 336/370 [01:29<00:09, 3.57it/s]
16065
  91%|█████████ | 337/370 [01:29<00:09, 3.51it/s]
16066
  91%|█████████▏| 338/370 [01:30<00:09, 3.39it/s]
16067
  92%|█████████▏| 339/370 [01:30<00:08, 3.46it/s]
16068
  92%|█████████▏| 340/370 [01:30<00:08, 3.36it/s]
16069
  92%|█████████▏| 341/370 [01:30<00:08, 3.28it/s]
16070
  92%|█████████▏| 342/370 [01:31<00:08, 3.43it/s]
16071
  93%|█████████▎| 343/370 [01:31<00:07, 3.40it/s]
16072
  93%|█████████▎| 344/370 [01:31<00:08, 3.09it/s]
16073
  93%|█████████▎| 345/370 [01:32<00:08, 2.87it/s]
16074
  94%|█████████▎| 346/370 [01:32<00:08, 2.90it/s]
16075
  94%|█████████▍| 347/370 [01:33<00:07, 2.88it/s]
16076
  94%|█████████▍| 348/370 [01:33<00:07, 3.03it/s]
16077
  94%|█████████▍| 349/370 [01:33<00:06, 3.06it/s]
16078
  95%|█████████▍| 350/370 [01:33<00:06, 3.18it/s]
16079
  95%|█████████▍| 351/370 [01:34<00:06, 3.12it/s]
16080
  95%|█████████▌| 352/370 [01:34<00:05, 3.06it/s]
16081
  95%|█████████▌| 353/370 [01:34<00:05, 3.08it/s]
16082
  96%|█████████▌| 354/370 [01:35<00:05, 3.01it/s]
16083
  96%|█████████▌| 355/370 [01:35<00:04, 3.24it/s]
16084
  96%|█████████▌| 356/370 [01:35<00:04, 3.43it/s]
16085
  96%|█████████▋| 357/370 [01:36<00:03, 3.58it/s]
16086
  97%|█████████▋| 358/370 [01:36<00:03, 3.88it/s]
16087
  97%|█████████▋| 359/370 [01:36<00:02, 3.93it/s]
16088
  97%|█████████▋| 360/370 [01:36<00:02, 3.61it/s]
16089
  98%|█████████▊| 361/370 [01:37<00:02, 3.70it/s]
16090
  98%|█████████▊| 362/370 [01:37<00:02, 3.78it/s]
16091
  98%|█████████▊| 363/370 [01:37<00:01, 3.73it/s]
16092
  98%|█████████▊| 364/370 [01:37<00:01, 3.56it/s]
16093
  99%|█████████▊| 365/370 [01:38<00:01, 3.81it/s]
16094
  99%|█████████▉| 366/370 [01:38<00:01, 3.57it/s]
16095
  99%|█████████▉| 367/370 [01:38<00:00, 3.69it/s]
16096
  99%|█████████▉| 368/370 [01:38<00:00, 3.73it/s]
16097
+ Saving model checkpoint to ./
16098
+ Configuration saved in ./config.json
16099
+ Model weights saved in ./pytorch_model.bin
16100
+ Configuration saved in ./preprocessor_config.json
runs/Feb06_11-36-42_job-a1cf84c8-7d28-46cd-9f0a-e09d16d4a1cb/events.out.tfevents.1644184605.job-a1cf84c8-7d28-46cd-9f0a-e09d16d4a1cb ADDED
@@ -0,0 +1,3 @@
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:355e5f5b1752478aec6d13be51f349e37a3b962618bccdf0cb7eec30b6f58784
3
+ size 40
train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 100.0,
3
+ "train_loss": 5.91964619928868,
4
+ "train_runtime": 36587.1788,
5
+ "train_samples": 6380,
6
+ "train_samples_per_second": 17.438,
7
+ "train_steps_per_second": 0.544
8
+ }
trainer_state.json ADDED
@@ -0,0 +1,1609 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 99.99749373433583,
5
+ "global_step": 19900,
6
+ "is_hyper_param_search": false,
7
+ "is_local_process_zero": true,
8
+ "is_world_process_zero": true,
9
+ "log_history": [
10
+ {
11
+ "epoch": 0.5,
12
+ "learning_rate": 3.6375e-06,
13
+ "loss": 136.5086,
14
+ "step": 100
15
+ },
16
+ {
17
+ "epoch": 1.01,
18
+ "learning_rate": 7.3875e-06,
19
+ "loss": 98.7644,
20
+ "step": 200
21
+ },
22
+ {
23
+ "epoch": 1.51,
24
+ "learning_rate": 1.1137499999999998e-05,
25
+ "loss": 81.5207,
26
+ "step": 300
27
+ },
28
+ {
29
+ "epoch": 2.01,
30
+ "learning_rate": 1.48875e-05,
31
+ "loss": 72.7656,
32
+ "step": 400
33
+ },
34
+ {
35
+ "epoch": 2.51,
36
+ "learning_rate": 1.86375e-05,
37
+ "loss": 64.6189,
38
+ "step": 500
39
+ },
40
+ {
41
+ "epoch": 2.51,
42
+ "eval_cer": 1.0,
43
+ "eval_loss": 63.80772399902344,
44
+ "eval_runtime": 134.6445,
45
+ "eval_samples_per_second": 21.969,
46
+ "eval_steps_per_second": 2.748,
47
+ "eval_wer": 1.0,
48
+ "step": 500
49
+ },
50
+ {
51
+ "epoch": 3.02,
52
+ "learning_rate": 2.23875e-05,
53
+ "loss": 53.9323,
54
+ "step": 600
55
+ },
56
+ {
57
+ "epoch": 3.52,
58
+ "learning_rate": 2.6137499999999995e-05,
59
+ "loss": 41.1416,
60
+ "step": 700
61
+ },
62
+ {
63
+ "epoch": 4.02,
64
+ "learning_rate": 2.9887499999999998e-05,
65
+ "loss": 27.1675,
66
+ "step": 800
67
+ },
68
+ {
69
+ "epoch": 4.52,
70
+ "learning_rate": 3.36375e-05,
71
+ "loss": 14.6933,
72
+ "step": 900
73
+ },
74
+ {
75
+ "epoch": 5.03,
76
+ "learning_rate": 3.7387499999999994e-05,
77
+ "loss": 8.0561,
78
+ "step": 1000
79
+ },
80
+ {
81
+ "epoch": 5.03,
82
+ "eval_cer": 1.0,
83
+ "eval_loss": 6.8014140129089355,
84
+ "eval_runtime": 123.7155,
85
+ "eval_samples_per_second": 23.91,
86
+ "eval_steps_per_second": 2.991,
87
+ "eval_wer": 1.0,
88
+ "step": 1000
89
+ },
90
+ {
91
+ "epoch": 5.53,
92
+ "learning_rate": 4.11375e-05,
93
+ "loss": 6.5122,
94
+ "step": 1100
95
+ },
96
+ {
97
+ "epoch": 6.03,
98
+ "learning_rate": 4.48875e-05,
99
+ "loss": 6.1973,
100
+ "step": 1200
101
+ },
102
+ {
103
+ "epoch": 6.53,
104
+ "learning_rate": 4.8637499999999996e-05,
105
+ "loss": 6.0687,
106
+ "step": 1300
107
+ },
108
+ {
109
+ "epoch": 7.04,
110
+ "learning_rate": 5.23875e-05,
111
+ "loss": 6.0963,
112
+ "step": 1400
113
+ },
114
+ {
115
+ "epoch": 7.54,
116
+ "learning_rate": 5.61375e-05,
117
+ "loss": 6.0427,
118
+ "step": 1500
119
+ },
120
+ {
121
+ "epoch": 7.54,
122
+ "eval_cer": 1.0,
123
+ "eval_loss": 6.074538230895996,
124
+ "eval_runtime": 122.7663,
125
+ "eval_samples_per_second": 24.095,
126
+ "eval_steps_per_second": 3.014,
127
+ "eval_wer": 1.0,
128
+ "step": 1500
129
+ },
130
+ {
131
+ "epoch": 8.04,
132
+ "learning_rate": 5.988749999999999e-05,
133
+ "loss": 6.0461,
134
+ "step": 1600
135
+ },
136
+ {
137
+ "epoch": 8.54,
138
+ "learning_rate": 6.36375e-05,
139
+ "loss": 5.9929,
140
+ "step": 1700
141
+ },
142
+ {
143
+ "epoch": 9.05,
144
+ "learning_rate": 6.738749999999999e-05,
145
+ "loss": 6.0191,
146
+ "step": 1800
147
+ },
148
+ {
149
+ "epoch": 9.55,
150
+ "learning_rate": 7.11375e-05,
151
+ "loss": 5.938,
152
+ "step": 1900
153
+ },
154
+ {
155
+ "epoch": 10.05,
156
+ "learning_rate": 7.48875e-05,
157
+ "loss": 5.9357,
158
+ "step": 2000
159
+ },
160
+ {
161
+ "epoch": 10.05,
162
+ "eval_cer": 1.0,
163
+ "eval_loss": 5.868249416351318,
164
+ "eval_runtime": 121.8577,
165
+ "eval_samples_per_second": 24.274,
166
+ "eval_steps_per_second": 3.036,
167
+ "eval_wer": 1.0,
168
+ "step": 2000
169
+ },
170
+ {
171
+ "epoch": 10.55,
172
+ "learning_rate": 7.45935754189944e-05,
173
+ "loss": 5.8143,
174
+ "step": 2100
175
+ },
176
+ {
177
+ "epoch": 11.06,
178
+ "learning_rate": 7.417458100558659e-05,
179
+ "loss": 5.7549,
180
+ "step": 2200
181
+ },
182
+ {
183
+ "epoch": 11.56,
184
+ "learning_rate": 7.375558659217876e-05,
185
+ "loss": 5.5454,
186
+ "step": 2300
187
+ },
188
+ {
189
+ "epoch": 12.06,
190
+ "learning_rate": 7.333659217877095e-05,
191
+ "loss": 5.3124,
192
+ "step": 2400
193
+ },
194
+ {
195
+ "epoch": 12.56,
196
+ "learning_rate": 7.291759776536312e-05,
197
+ "loss": 5.0489,
198
+ "step": 2500
199
+ },
200
+ {
201
+ "epoch": 12.56,
202
+ "eval_cer": 0.7749867254829882,
203
+ "eval_loss": 4.4031829833984375,
204
+ "eval_runtime": 123.6444,
205
+ "eval_samples_per_second": 23.923,
206
+ "eval_steps_per_second": 2.992,
207
+ "eval_wer": 0.9989858012170385,
208
+ "step": 2500
209
+ },
210
+ {
211
+ "epoch": 13.07,
212
+ "learning_rate": 7.249860335195531e-05,
213
+ "loss": 5.0055,
214
+ "step": 2600
215
+ },
216
+ {
217
+ "epoch": 13.57,
218
+ "learning_rate": 7.207960893854748e-05,
219
+ "loss": 4.8755,
220
+ "step": 2700
221
+ },
222
+ {
223
+ "epoch": 14.07,
224
+ "learning_rate": 7.166480446927373e-05,
225
+ "loss": 4.7729,
226
+ "step": 2800
227
+ },
228
+ {
229
+ "epoch": 14.57,
230
+ "learning_rate": 7.124581005586592e-05,
231
+ "loss": 4.7114,
232
+ "step": 2900
233
+ },
234
+ {
235
+ "epoch": 15.08,
236
+ "learning_rate": 7.083100558659217e-05,
237
+ "loss": 4.6184,
238
+ "step": 3000
239
+ },
240
+ {
241
+ "epoch": 15.08,
242
+ "eval_cer": 0.6767552995956377,
243
+ "eval_loss": 3.83825421333313,
244
+ "eval_runtime": 124.4317,
245
+ "eval_samples_per_second": 23.772,
246
+ "eval_steps_per_second": 2.974,
247
+ "eval_wer": 0.9983096686950642,
248
+ "step": 3000
249
+ },
250
+ {
251
+ "epoch": 15.58,
252
+ "learning_rate": 7.041201117318434e-05,
253
+ "loss": 4.5657,
254
+ "step": 3100
255
+ },
256
+ {
257
+ "epoch": 16.08,
258
+ "learning_rate": 6.999301675977653e-05,
259
+ "loss": 4.5613,
260
+ "step": 3200
261
+ },
262
+ {
263
+ "epoch": 16.58,
264
+ "learning_rate": 6.95782122905028e-05,
265
+ "loss": 4.4679,
266
+ "step": 3300
267
+ },
268
+ {
269
+ "epoch": 17.09,
270
+ "learning_rate": 6.915921787709497e-05,
271
+ "loss": 4.4288,
272
+ "step": 3400
273
+ },
274
+ {
275
+ "epoch": 17.59,
276
+ "learning_rate": 6.874022346368715e-05,
277
+ "loss": 4.365,
278
+ "step": 3500
279
+ },
280
+ {
281
+ "epoch": 17.59,
282
+ "eval_cer": 0.62994731037863,
283
+ "eval_loss": 3.4632537364959717,
284
+ "eval_runtime": 122.6998,
285
+ "eval_samples_per_second": 24.108,
286
+ "eval_steps_per_second": 3.015,
287
+ "eval_wer": 0.9959432048681541,
288
+ "step": 3500
289
+ },
290
+ {
291
+ "epoch": 18.09,
292
+ "learning_rate": 6.832122905027933e-05,
293
+ "loss": 4.3607,
294
+ "step": 3600
295
+ },
296
+ {
297
+ "epoch": 18.59,
298
+ "learning_rate": 6.790223463687151e-05,
299
+ "loss": 4.2659,
300
+ "step": 3700
301
+ },
302
+ {
303
+ "epoch": 19.1,
304
+ "learning_rate": 6.748324022346368e-05,
305
+ "loss": 4.2805,
306
+ "step": 3800
307
+ },
308
+ {
309
+ "epoch": 19.6,
310
+ "learning_rate": 6.706424581005587e-05,
311
+ "loss": 4.1783,
312
+ "step": 3900
313
+ },
314
+ {
315
+ "epoch": 20.1,
316
+ "learning_rate": 6.664525139664804e-05,
317
+ "loss": 4.1026,
318
+ "step": 4000
319
+ },
320
+ {
321
+ "epoch": 20.1,
322
+ "eval_cer": 0.5813830004492914,
323
+ "eval_loss": 3.073154926300049,
324
+ "eval_runtime": 122.4373,
325
+ "eval_samples_per_second": 24.159,
326
+ "eval_steps_per_second": 3.022,
327
+ "eval_wer": 0.9901960784313726,
328
+ "step": 4000
329
+ },
330
+ {
331
+ "epoch": 20.6,
332
+ "learning_rate": 6.622625698324022e-05,
333
+ "loss": 4.0481,
334
+ "step": 4100
335
+ },
336
+ {
337
+ "epoch": 21.11,
338
+ "learning_rate": 6.58072625698324e-05,
339
+ "loss": 4.005,
340
+ "step": 4200
341
+ },
342
+ {
343
+ "epoch": 21.61,
344
+ "learning_rate": 6.538826815642457e-05,
345
+ "loss": 3.966,
346
+ "step": 4300
347
+ },
348
+ {
349
+ "epoch": 22.11,
350
+ "learning_rate": 6.496927374301676e-05,
351
+ "loss": 3.9327,
352
+ "step": 4400
353
+ },
354
+ {
355
+ "epoch": 22.61,
356
+ "learning_rate": 6.455027932960893e-05,
357
+ "loss": 3.8655,
358
+ "step": 4500
359
+ },
360
+ {
361
+ "epoch": 22.61,
362
+ "eval_cer": 0.5465016542090431,
363
+ "eval_loss": 2.7638278007507324,
364
+ "eval_runtime": 121.1871,
365
+ "eval_samples_per_second": 24.409,
366
+ "eval_steps_per_second": 3.053,
367
+ "eval_wer": 0.986815415821501,
368
+ "step": 4500
369
+ },
370
+ {
371
+ "epoch": 23.12,
372
+ "learning_rate": 6.413128491620112e-05,
373
+ "loss": 3.8466,
374
+ "step": 4600
375
+ },
376
+ {
377
+ "epoch": 23.62,
378
+ "learning_rate": 6.371229050279329e-05,
379
+ "loss": 3.7785,
380
+ "step": 4700
381
+ },
382
+ {
383
+ "epoch": 24.12,
384
+ "learning_rate": 6.329329608938548e-05,
385
+ "loss": 3.7262,
386
+ "step": 4800
387
+ },
388
+ {
389
+ "epoch": 24.62,
390
+ "learning_rate": 6.287430167597765e-05,
391
+ "loss": 3.6768,
392
+ "step": 4900
393
+ },
394
+ {
395
+ "epoch": 25.13,
396
+ "learning_rate": 6.245530726256982e-05,
397
+ "loss": 3.6991,
398
+ "step": 5000
399
+ },
400
+ {
401
+ "epoch": 25.13,
402
+ "eval_cer": 0.5088020258955194,
403
+ "eval_loss": 2.475937604904175,
404
+ "eval_runtime": 122.6664,
405
+ "eval_samples_per_second": 24.114,
406
+ "eval_steps_per_second": 3.016,
407
+ "eval_wer": 0.9810682893847193,
408
+ "step": 5000
409
+ },
410
+ {
411
+ "epoch": 25.63,
412
+ "learning_rate": 6.203631284916201e-05,
413
+ "loss": 3.5971,
414
+ "step": 5100
415
+ },
416
+ {
417
+ "epoch": 26.13,
418
+ "learning_rate": 6.161731843575418e-05,
419
+ "loss": 3.6243,
420
+ "step": 5200
421
+ },
422
+ {
423
+ "epoch": 26.63,
424
+ "learning_rate": 6.119832402234637e-05,
425
+ "loss": 3.5803,
426
+ "step": 5300
427
+ },
428
+ {
429
+ "epoch": 27.14,
430
+ "learning_rate": 6.077932960893854e-05,
431
+ "loss": 3.5697,
432
+ "step": 5400
433
+ },
434
+ {
435
+ "epoch": 27.64,
436
+ "learning_rate": 6.03645251396648e-05,
437
+ "loss": 3.4894,
438
+ "step": 5500
439
+ },
440
+ {
441
+ "epoch": 27.64,
442
+ "eval_cer": 0.4851529632806437,
443
+ "eval_loss": 2.2937276363372803,
444
+ "eval_runtime": 122.0039,
445
+ "eval_samples_per_second": 24.245,
446
+ "eval_steps_per_second": 3.033,
447
+ "eval_wer": 0.9746450304259635,
448
+ "step": 5500
449
+ },
450
+ {
451
+ "epoch": 28.14,
452
+ "learning_rate": 5.994553072625698e-05,
453
+ "loss": 3.5363,
454
+ "step": 5600
455
+ },
456
+ {
457
+ "epoch": 28.64,
458
+ "learning_rate": 5.952653631284916e-05,
459
+ "loss": 3.4932,
460
+ "step": 5700
461
+ },
462
+ {
463
+ "epoch": 29.15,
464
+ "learning_rate": 5.910754189944134e-05,
465
+ "loss": 3.436,
466
+ "step": 5800
467
+ },
468
+ {
469
+ "epoch": 29.65,
470
+ "learning_rate": 5.868854748603351e-05,
471
+ "loss": 3.4209,
472
+ "step": 5900
473
+ },
474
+ {
475
+ "epoch": 30.15,
476
+ "learning_rate": 5.8269553072625696e-05,
477
+ "loss": 3.3983,
478
+ "step": 6000
479
+ },
480
+ {
481
+ "epoch": 30.15,
482
+ "eval_cer": 0.4673855328186905,
483
+ "eval_loss": 2.168361186981201,
484
+ "eval_runtime": 121.2096,
485
+ "eval_samples_per_second": 24.404,
486
+ "eval_steps_per_second": 3.053,
487
+ "eval_wer": 0.9732927653820149,
488
+ "step": 6000
489
+ },
490
+ {
491
+ "epoch": 30.65,
492
+ "learning_rate": 5.785055865921787e-05,
493
+ "loss": 3.3573,
494
+ "step": 6100
495
+ },
496
+ {
497
+ "epoch": 31.16,
498
+ "learning_rate": 5.743156424581005e-05,
499
+ "loss": 3.3821,
500
+ "step": 6200
501
+ },
502
+ {
503
+ "epoch": 31.66,
504
+ "learning_rate": 5.701256983240223e-05,
505
+ "loss": 3.3188,
506
+ "step": 6300
507
+ },
508
+ {
509
+ "epoch": 32.16,
510
+ "learning_rate": 5.659357541899441e-05,
511
+ "loss": 3.3519,
512
+ "step": 6400
513
+ },
514
+ {
515
+ "epoch": 32.66,
516
+ "learning_rate": 5.6174581005586586e-05,
517
+ "loss": 3.2736,
518
+ "step": 6500
519
+ },
520
+ {
521
+ "epoch": 32.66,
522
+ "eval_cer": 0.44581954825797493,
523
+ "eval_loss": 2.037248134613037,
524
+ "eval_runtime": 123.5198,
525
+ "eval_samples_per_second": 23.948,
526
+ "eval_steps_per_second": 2.995,
527
+ "eval_wer": 0.9658553076402975,
528
+ "step": 6500
529
+ },
530
+ {
531
+ "epoch": 33.17,
532
+ "learning_rate": 5.5755586592178765e-05,
533
+ "loss": 3.3093,
534
+ "step": 6600
535
+ },
536
+ {
537
+ "epoch": 33.67,
538
+ "learning_rate": 5.5336592178770945e-05,
539
+ "loss": 3.2599,
540
+ "step": 6700
541
+ },
542
+ {
543
+ "epoch": 34.17,
544
+ "learning_rate": 5.4917597765363124e-05,
545
+ "loss": 3.2115,
546
+ "step": 6800
547
+ },
548
+ {
549
+ "epoch": 34.67,
550
+ "learning_rate": 5.4498603351955304e-05,
551
+ "loss": 3.1972,
552
+ "step": 6900
553
+ },
554
+ {
555
+ "epoch": 35.18,
556
+ "learning_rate": 5.407960893854748e-05,
557
+ "loss": 3.1884,
558
+ "step": 7000
559
+ },
560
+ {
561
+ "epoch": 35.18,
562
+ "eval_cer": 0.4329126332557285,
563
+ "eval_loss": 1.9266635179519653,
564
+ "eval_runtime": 122.7815,
565
+ "eval_samples_per_second": 24.092,
566
+ "eval_steps_per_second": 3.013,
567
+ "eval_wer": 0.964841108857336,
568
+ "step": 7000
569
+ },
570
+ {
571
+ "epoch": 35.68,
572
+ "learning_rate": 5.366061452513966e-05,
573
+ "loss": 3.1908,
574
+ "step": 7100
575
+ },
576
+ {
577
+ "epoch": 36.18,
578
+ "learning_rate": 5.324162011173184e-05,
579
+ "loss": 3.1505,
580
+ "step": 7200
581
+ },
582
+ {
583
+ "epoch": 36.68,
584
+ "learning_rate": 5.2822625698324014e-05,
585
+ "loss": 3.134,
586
+ "step": 7300
587
+ },
588
+ {
589
+ "epoch": 37.19,
590
+ "learning_rate": 5.24036312849162e-05,
591
+ "loss": 3.101,
592
+ "step": 7400
593
+ },
594
+ {
595
+ "epoch": 37.69,
596
+ "learning_rate": 5.198882681564245e-05,
597
+ "loss": 3.1248,
598
+ "step": 7500
599
+ },
600
+ {
601
+ "epoch": 37.69,
602
+ "eval_cer": 0.4217211942980844,
603
+ "eval_loss": 1.8408104181289673,
604
+ "eval_runtime": 122.5199,
605
+ "eval_samples_per_second": 24.143,
606
+ "eval_steps_per_second": 3.02,
607
+ "eval_wer": 0.9590939824205544,
608
+ "step": 7500
609
+ },
610
+ {
611
+ "epoch": 38.19,
612
+ "learning_rate": 5.156983240223463e-05,
613
+ "loss": 3.0958,
614
+ "step": 7600
615
+ },
616
+ {
617
+ "epoch": 38.69,
618
+ "learning_rate": 5.115083798882681e-05,
619
+ "loss": 3.0627,
620
+ "step": 7700
621
+ },
622
+ {
623
+ "epoch": 39.2,
624
+ "learning_rate": 5.073184357541899e-05,
625
+ "loss": 3.0716,
626
+ "step": 7800
627
+ },
628
+ {
629
+ "epoch": 39.7,
630
+ "learning_rate": 5.031284916201117e-05,
631
+ "loss": 3.0455,
632
+ "step": 7900
633
+ },
634
+ {
635
+ "epoch": 40.2,
636
+ "learning_rate": 4.989804469273743e-05,
637
+ "loss": 3.0381,
638
+ "step": 8000
639
+ },
640
+ {
641
+ "epoch": 40.2,
642
+ "eval_cer": 0.40742556059306456,
643
+ "eval_loss": 1.7530696392059326,
644
+ "eval_runtime": 125.5769,
645
+ "eval_samples_per_second": 23.555,
646
+ "eval_steps_per_second": 2.946,
647
+ "eval_wer": 0.9503042596348884,
648
+ "step": 8000
649
+ },
650
+ {
651
+ "epoch": 40.7,
652
+ "learning_rate": 4.94790502793296e-05,
653
+ "loss": 2.9917,
654
+ "step": 8100
655
+ },
656
+ {
657
+ "epoch": 41.21,
658
+ "learning_rate": 4.906005586592179e-05,
659
+ "loss": 3.0649,
660
+ "step": 8200
661
+ },
662
+ {
663
+ "epoch": 41.71,
664
+ "learning_rate": 4.864106145251396e-05,
665
+ "loss": 2.9746,
666
+ "step": 8300
667
+ },
668
+ {
669
+ "epoch": 42.21,
670
+ "learning_rate": 4.8222067039106146e-05,
671
+ "loss": 2.9459,
672
+ "step": 8400
673
+ },
674
+ {
675
+ "epoch": 42.71,
676
+ "learning_rate": 4.780307262569832e-05,
677
+ "loss": 2.9515,
678
+ "step": 8500
679
+ },
680
+ {
681
+ "epoch": 42.71,
682
+ "eval_cer": 0.39672425764816405,
683
+ "eval_loss": 1.6879578828811646,
684
+ "eval_runtime": 121.7673,
685
+ "eval_samples_per_second": 24.292,
686
+ "eval_steps_per_second": 3.039,
687
+ "eval_wer": 0.9459093982420554,
688
+ "step": 8500
689
+ },
690
+ {
691
+ "epoch": 43.22,
692
+ "learning_rate": 4.7384078212290505e-05,
693
+ "loss": 2.9478,
694
+ "step": 8600
695
+ },
696
+ {
697
+ "epoch": 43.72,
698
+ "learning_rate": 4.696508379888268e-05,
699
+ "loss": 2.9046,
700
+ "step": 8700
701
+ },
702
+ {
703
+ "epoch": 44.22,
704
+ "learning_rate": 4.654608938547485e-05,
705
+ "loss": 2.9345,
706
+ "step": 8800
707
+ },
708
+ {
709
+ "epoch": 44.72,
710
+ "learning_rate": 4.6127094972067036e-05,
711
+ "loss": 2.8722,
712
+ "step": 8900
713
+ },
714
+ {
715
+ "epoch": 45.23,
716
+ "learning_rate": 4.570810055865921e-05,
717
+ "loss": 2.8704,
718
+ "step": 9000
719
+ },
720
+ {
721
+ "epoch": 45.23,
722
+ "eval_cer": 0.3884327900992525,
723
+ "eval_loss": 1.626428484916687,
724
+ "eval_runtime": 124.9171,
725
+ "eval_samples_per_second": 23.68,
726
+ "eval_steps_per_second": 2.962,
727
+ "eval_wer": 0.9377958079783637,
728
+ "step": 9000
729
+ },
730
+ {
731
+ "epoch": 45.73,
732
+ "learning_rate": 4.5289106145251395e-05,
733
+ "loss": 2.8807,
734
+ "step": 9100
735
+ },
736
+ {
737
+ "epoch": 46.23,
738
+ "learning_rate": 4.487011173184357e-05,
739
+ "loss": 2.8631,
740
+ "step": 9200
741
+ },
742
+ {
743
+ "epoch": 46.73,
744
+ "learning_rate": 4.4451117318435753e-05,
745
+ "loss": 2.8743,
746
+ "step": 9300
747
+ },
748
+ {
749
+ "epoch": 47.24,
750
+ "learning_rate": 4.4032122905027926e-05,
751
+ "loss": 2.8364,
752
+ "step": 9400
753
+ },
754
+ {
755
+ "epoch": 47.74,
756
+ "learning_rate": 4.361312849162011e-05,
757
+ "loss": 2.8128,
758
+ "step": 9500
759
+ },
760
+ {
761
+ "epoch": 47.74,
762
+ "eval_cer": 0.37818077849936693,
763
+ "eval_loss": 1.5620697736740112,
764
+ "eval_runtime": 128.811,
765
+ "eval_samples_per_second": 22.964,
766
+ "eval_steps_per_second": 2.872,
767
+ "eval_wer": 0.934077079107505,
768
+ "step": 9500
769
+ },
770
+ {
771
+ "epoch": 48.24,
772
+ "learning_rate": 4.3194134078212285e-05,
773
+ "loss": 2.855,
774
+ "step": 9600
775
+ },
776
+ {
777
+ "epoch": 48.74,
778
+ "learning_rate": 4.2775139664804464e-05,
779
+ "loss": 2.7587,
780
+ "step": 9700
781
+ },
782
+ {
783
+ "epoch": 49.25,
784
+ "learning_rate": 4.2356145251396644e-05,
785
+ "loss": 2.7794,
786
+ "step": 9800
787
+ },
788
+ {
789
+ "epoch": 49.75,
790
+ "learning_rate": 4.193715083798882e-05,
791
+ "loss": 2.7664,
792
+ "step": 9900
793
+ },
794
+ {
795
+ "epoch": 50.25,
796
+ "learning_rate": 4.1518156424581e-05,
797
+ "loss": 2.7386,
798
+ "step": 10000
799
+ },
800
+ {
801
+ "epoch": 50.25,
802
+ "eval_cer": 0.3663766695257934,
803
+ "eval_loss": 1.5010998249053955,
804
+ "eval_runtime": 124.8062,
805
+ "eval_samples_per_second": 23.701,
806
+ "eval_steps_per_second": 2.965,
807
+ "eval_wer": 0.9242731575388776,
808
+ "step": 10000
809
+ },
810
+ {
811
+ "epoch": 50.75,
812
+ "learning_rate": 4.109916201117318e-05,
813
+ "loss": 2.7295,
814
+ "step": 10100
815
+ },
816
+ {
817
+ "epoch": 51.26,
818
+ "learning_rate": 4.0684357541899433e-05,
819
+ "loss": 2.7372,
820
+ "step": 10200
821
+ },
822
+ {
823
+ "epoch": 51.76,
824
+ "learning_rate": 4.026536312849162e-05,
825
+ "loss": 2.7116,
826
+ "step": 10300
827
+ },
828
+ {
829
+ "epoch": 52.26,
830
+ "learning_rate": 3.984636871508379e-05,
831
+ "loss": 2.6882,
832
+ "step": 10400
833
+ },
834
+ {
835
+ "epoch": 52.76,
836
+ "learning_rate": 3.942737430167598e-05,
837
+ "loss": 2.6646,
838
+ "step": 10500
839
+ },
840
+ {
841
+ "epoch": 52.76,
842
+ "eval_cer": 0.3574725319609525,
843
+ "eval_loss": 1.460774540901184,
844
+ "eval_runtime": 125.7182,
845
+ "eval_samples_per_second": 23.529,
846
+ "eval_steps_per_second": 2.943,
847
+ "eval_wer": 0.9192021636240704,
848
+ "step": 10500
849
+ },
850
+ {
851
+ "epoch": 53.27,
852
+ "learning_rate": 3.900837988826815e-05,
853
+ "loss": 2.6904,
854
+ "step": 10600
855
+ },
856
+ {
857
+ "epoch": 53.77,
858
+ "learning_rate": 3.858938547486034e-05,
859
+ "loss": 2.6589,
860
+ "step": 10700
861
+ },
862
+ {
863
+ "epoch": 54.27,
864
+ "learning_rate": 3.817039106145251e-05,
865
+ "loss": 2.6618,
866
+ "step": 10800
867
+ },
868
+ {
869
+ "epoch": 54.77,
870
+ "learning_rate": 3.7751396648044696e-05,
871
+ "loss": 2.6187,
872
+ "step": 10900
873
+ },
874
+ {
875
+ "epoch": 55.28,
876
+ "learning_rate": 3.733240223463687e-05,
877
+ "loss": 2.6072,
878
+ "step": 11000
879
+ },
880
+ {
881
+ "epoch": 55.28,
882
+ "eval_cer": 0.3501204917697995,
883
+ "eval_loss": 1.4251333475112915,
884
+ "eval_runtime": 122.3324,
885
+ "eval_samples_per_second": 24.18,
886
+ "eval_steps_per_second": 3.025,
887
+ "eval_wer": 0.9148073022312373,
888
+ "step": 11000
889
+ },
890
+ {
891
+ "epoch": 55.78,
892
+ "learning_rate": 3.691340782122905e-05,
893
+ "loss": 2.6307,
894
+ "step": 11100
895
+ },
896
+ {
897
+ "epoch": 56.28,
898
+ "learning_rate": 3.649441340782123e-05,
899
+ "loss": 2.5884,
900
+ "step": 11200
901
+ },
902
+ {
903
+ "epoch": 56.78,
904
+ "learning_rate": 3.6075418994413407e-05,
905
+ "loss": 2.5678,
906
+ "step": 11300
907
+ },
908
+ {
909
+ "epoch": 57.29,
910
+ "learning_rate": 3.5656424581005586e-05,
911
+ "loss": 2.5722,
912
+ "step": 11400
913
+ },
914
+ {
915
+ "epoch": 57.79,
916
+ "learning_rate": 3.523743016759776e-05,
917
+ "loss": 2.569,
918
+ "step": 11500
919
+ },
920
+ {
921
+ "epoch": 57.79,
922
+ "eval_cer": 0.34615855900012255,
923
+ "eval_loss": 1.3836983442306519,
924
+ "eval_runtime": 125.4562,
925
+ "eval_samples_per_second": 23.578,
926
+ "eval_steps_per_second": 2.949,
927
+ "eval_wer": 0.9060175794455714,
928
+ "step": 11500
929
+ },
930
+ {
931
+ "epoch": 58.29,
932
+ "learning_rate": 3.481843575418994e-05,
933
+ "loss": 2.5487,
934
+ "step": 11600
935
+ },
936
+ {
937
+ "epoch": 58.79,
938
+ "learning_rate": 3.439944134078212e-05,
939
+ "loss": 2.5536,
940
+ "step": 11700
941
+ },
942
+ {
943
+ "epoch": 59.3,
944
+ "learning_rate": 3.39804469273743e-05,
945
+ "loss": 2.5402,
946
+ "step": 11800
947
+ },
948
+ {
949
+ "epoch": 59.8,
950
+ "learning_rate": 3.3561452513966476e-05,
951
+ "loss": 2.5001,
952
+ "step": 11900
953
+ },
954
+ {
955
+ "epoch": 60.3,
956
+ "learning_rate": 3.3142458100558655e-05,
957
+ "loss": 2.5091,
958
+ "step": 12000
959
+ },
960
+ {
961
+ "epoch": 60.3,
962
+ "eval_cer": 0.3392149654862558,
963
+ "eval_loss": 1.3589394092559814,
964
+ "eval_runtime": 123.5116,
965
+ "eval_samples_per_second": 23.949,
966
+ "eval_steps_per_second": 2.996,
967
+ "eval_wer": 0.9070317782285328,
968
+ "step": 12000
969
+ },
970
+ {
971
+ "epoch": 60.8,
972
+ "learning_rate": 3.2723463687150835e-05,
973
+ "loss": 2.4874,
974
+ "step": 12100
975
+ },
976
+ {
977
+ "epoch": 61.31,
978
+ "learning_rate": 3.2304469273743014e-05,
979
+ "loss": 2.4834,
980
+ "step": 12200
981
+ },
982
+ {
983
+ "epoch": 61.81,
984
+ "learning_rate": 3.1885474860335194e-05,
985
+ "loss": 2.478,
986
+ "step": 12300
987
+ },
988
+ {
989
+ "epoch": 62.31,
990
+ "learning_rate": 3.147067039106145e-05,
991
+ "loss": 2.4811,
992
+ "step": 12400
993
+ },
994
+ {
995
+ "epoch": 62.81,
996
+ "learning_rate": 3.105167597765363e-05,
997
+ "loss": 2.4588,
998
+ "step": 12500
999
+ },
1000
+ {
1001
+ "epoch": 62.81,
1002
+ "eval_cer": 0.3283911285381693,
1003
+ "eval_loss": 1.326094627380371,
1004
+ "eval_runtime": 126.176,
1005
+ "eval_samples_per_second": 23.443,
1006
+ "eval_steps_per_second": 2.932,
1007
+ "eval_wer": 0.896551724137931,
1008
+ "step": 12500
1009
+ },
1010
+ {
1011
+ "epoch": 63.32,
1012
+ "learning_rate": 3.063268156424581e-05,
1013
+ "loss": 2.4296,
1014
+ "step": 12600
1015
+ },
1016
+ {
1017
+ "epoch": 63.82,
1018
+ "learning_rate": 3.0213687150837987e-05,
1019
+ "loss": 2.4535,
1020
+ "step": 12700
1021
+ },
1022
+ {
1023
+ "epoch": 64.32,
1024
+ "learning_rate": 2.9794692737430163e-05,
1025
+ "loss": 2.4572,
1026
+ "step": 12800
1027
+ },
1028
+ {
1029
+ "epoch": 64.82,
1030
+ "learning_rate": 2.9375698324022342e-05,
1031
+ "loss": 2.3824,
1032
+ "step": 12900
1033
+ },
1034
+ {
1035
+ "epoch": 65.33,
1036
+ "learning_rate": 2.896508379888268e-05,
1037
+ "loss": 2.4083,
1038
+ "step": 13000
1039
+ },
1040
+ {
1041
+ "epoch": 65.33,
1042
+ "eval_cer": 0.32647142915492383,
1043
+ "eval_loss": 1.3052246570587158,
1044
+ "eval_runtime": 125.7147,
1045
+ "eval_samples_per_second": 23.529,
1046
+ "eval_steps_per_second": 2.943,
1047
+ "eval_wer": 0.8982420554428668,
1048
+ "step": 13000
1049
+ },
1050
+ {
1051
+ "epoch": 65.83,
1052
+ "learning_rate": 2.854608938547486e-05,
1053
+ "loss": 2.4015,
1054
+ "step": 13100
1055
+ },
1056
+ {
1057
+ "epoch": 66.33,
1058
+ "learning_rate": 2.812709497206704e-05,
1059
+ "loss": 2.3747,
1060
+ "step": 13200
1061
+ },
1062
+ {
1063
+ "epoch": 66.83,
1064
+ "learning_rate": 2.7708100558659218e-05,
1065
+ "loss": 2.3625,
1066
+ "step": 13300
1067
+ },
1068
+ {
1069
+ "epoch": 67.34,
1070
+ "learning_rate": 2.728910614525139e-05,
1071
+ "loss": 2.4027,
1072
+ "step": 13400
1073
+ },
1074
+ {
1075
+ "epoch": 67.84,
1076
+ "learning_rate": 2.687011173184357e-05,
1077
+ "loss": 2.3787,
1078
+ "step": 13500
1079
+ },
1080
+ {
1081
+ "epoch": 67.84,
1082
+ "eval_cer": 0.32430666176530654,
1083
+ "eval_loss": 1.2997361421585083,
1084
+ "eval_runtime": 122.085,
1085
+ "eval_samples_per_second": 24.229,
1086
+ "eval_steps_per_second": 3.031,
1087
+ "eval_wer": 0.8908045977011494,
1088
+ "step": 13500
1089
+ },
1090
+ {
1091
+ "epoch": 68.34,
1092
+ "learning_rate": 2.645111731843575e-05,
1093
+ "loss": 2.3351,
1094
+ "step": 13600
1095
+ },
1096
+ {
1097
+ "epoch": 68.84,
1098
+ "learning_rate": 2.603212290502793e-05,
1099
+ "loss": 2.3445,
1100
+ "step": 13700
1101
+ },
1102
+ {
1103
+ "epoch": 69.35,
1104
+ "learning_rate": 2.561312849162011e-05,
1105
+ "loss": 2.3649,
1106
+ "step": 13800
1107
+ },
1108
+ {
1109
+ "epoch": 69.85,
1110
+ "learning_rate": 2.5194134078212288e-05,
1111
+ "loss": 2.326,
1112
+ "step": 13900
1113
+ },
1114
+ {
1115
+ "epoch": 70.35,
1116
+ "learning_rate": 2.4775139664804467e-05,
1117
+ "loss": 2.3457,
1118
+ "step": 14000
1119
+ },
1120
+ {
1121
+ "epoch": 70.35,
1122
+ "eval_cer": 0.3187109422864845,
1123
+ "eval_loss": 1.2778037786483765,
1124
+ "eval_runtime": 128.0452,
1125
+ "eval_samples_per_second": 23.101,
1126
+ "eval_steps_per_second": 2.89,
1127
+ "eval_wer": 0.889790398918188,
1128
+ "step": 14000
1129
+ },
1130
+ {
1131
+ "epoch": 70.85,
1132
+ "learning_rate": 2.4356145251396646e-05,
1133
+ "loss": 2.2999,
1134
+ "step": 14100
1135
+ },
1136
+ {
1137
+ "epoch": 71.36,
1138
+ "learning_rate": 2.3937150837988826e-05,
1139
+ "loss": 2.3383,
1140
+ "step": 14200
1141
+ },
1142
+ {
1143
+ "epoch": 71.86,
1144
+ "learning_rate": 2.3518156424581005e-05,
1145
+ "loss": 2.312,
1146
+ "step": 14300
1147
+ },
1148
+ {
1149
+ "epoch": 72.36,
1150
+ "learning_rate": 2.309916201117318e-05,
1151
+ "loss": 2.296,
1152
+ "step": 14400
1153
+ },
1154
+ {
1155
+ "epoch": 72.86,
1156
+ "learning_rate": 2.268016759776536e-05,
1157
+ "loss": 2.3099,
1158
+ "step": 14500
1159
+ },
1160
+ {
1161
+ "epoch": 72.86,
1162
+ "eval_cer": 0.3172405342482539,
1163
+ "eval_loss": 1.266068935394287,
1164
+ "eval_runtime": 123.4423,
1165
+ "eval_samples_per_second": 23.963,
1166
+ "eval_steps_per_second": 2.997,
1167
+ "eval_wer": 0.8830290736984449,
1168
+ "step": 14500
1169
+ },
1170
+ {
1171
+ "epoch": 73.37,
1172
+ "learning_rate": 2.226117318435754e-05,
1173
+ "loss": 2.2976,
1174
+ "step": 14600
1175
+ },
1176
+ {
1177
+ "epoch": 73.87,
1178
+ "learning_rate": 2.184217877094972e-05,
1179
+ "loss": 2.3002,
1180
+ "step": 14700
1181
+ },
1182
+ {
1183
+ "epoch": 74.37,
1184
+ "learning_rate": 2.14231843575419e-05,
1185
+ "loss": 2.2536,
1186
+ "step": 14800
1187
+ },
1188
+ {
1189
+ "epoch": 74.87,
1190
+ "learning_rate": 2.1004189944134078e-05,
1191
+ "loss": 2.2683,
1192
+ "step": 14900
1193
+ },
1194
+ {
1195
+ "epoch": 75.38,
1196
+ "learning_rate": 2.0585195530726257e-05,
1197
+ "loss": 2.2559,
1198
+ "step": 15000
1199
+ },
1200
+ {
1201
+ "epoch": 75.38,
1202
+ "eval_cer": 0.3143405628395213,
1203
+ "eval_loss": 1.2474771738052368,
1204
+ "eval_runtime": 124.7835,
1205
+ "eval_samples_per_second": 23.705,
1206
+ "eval_steps_per_second": 2.965,
1207
+ "eval_wer": 0.8850574712643678,
1208
+ "step": 15000
1209
+ },
1210
+ {
1211
+ "epoch": 75.88,
1212
+ "learning_rate": 2.0166201117318437e-05,
1213
+ "loss": 2.2334,
1214
+ "step": 15100
1215
+ },
1216
+ {
1217
+ "epoch": 76.38,
1218
+ "learning_rate": 1.9747206703910616e-05,
1219
+ "loss": 2.2153,
1220
+ "step": 15200
1221
+ },
1222
+ {
1223
+ "epoch": 76.88,
1224
+ "learning_rate": 1.932821229050279e-05,
1225
+ "loss": 2.2244,
1226
+ "step": 15300
1227
+ },
1228
+ {
1229
+ "epoch": 77.39,
1230
+ "learning_rate": 1.8909217877094968e-05,
1231
+ "loss": 2.2327,
1232
+ "step": 15400
1233
+ },
1234
+ {
1235
+ "epoch": 77.89,
1236
+ "learning_rate": 1.8490223463687148e-05,
1237
+ "loss": 2.2264,
1238
+ "step": 15500
1239
+ },
1240
+ {
1241
+ "epoch": 77.89,
1242
+ "eval_cer": 0.3085406200220561,
1243
+ "eval_loss": 1.2318875789642334,
1244
+ "eval_runtime": 127.717,
1245
+ "eval_samples_per_second": 23.161,
1246
+ "eval_steps_per_second": 2.897,
1247
+ "eval_wer": 0.8739012846517917,
1248
+ "step": 15500
1249
+ },
1250
+ {
1251
+ "epoch": 78.39,
1252
+ "learning_rate": 1.8071229050279327e-05,
1253
+ "loss": 2.2722,
1254
+ "step": 15600
1255
+ },
1256
+ {
1257
+ "epoch": 78.89,
1258
+ "learning_rate": 1.7652234636871506e-05,
1259
+ "loss": 2.1997,
1260
+ "step": 15700
1261
+ },
1262
+ {
1263
+ "epoch": 79.4,
1264
+ "learning_rate": 1.7233240223463686e-05,
1265
+ "loss": 2.1979,
1266
+ "step": 15800
1267
+ },
1268
+ {
1269
+ "epoch": 79.9,
1270
+ "learning_rate": 1.6814245810055865e-05,
1271
+ "loss": 2.222,
1272
+ "step": 15900
1273
+ },
1274
+ {
1275
+ "epoch": 80.4,
1276
+ "learning_rate": 1.639525139664804e-05,
1277
+ "loss": 2.196,
1278
+ "step": 16000
1279
+ },
1280
+ {
1281
+ "epoch": 80.4,
1282
+ "eval_cer": 0.3048645999264796,
1283
+ "eval_loss": 1.2218027114868164,
1284
+ "eval_runtime": 124.9046,
1285
+ "eval_samples_per_second": 23.682,
1286
+ "eval_steps_per_second": 2.962,
1287
+ "eval_wer": 0.8722109533468559,
1288
+ "step": 16000
1289
+ },
1290
+ {
1291
+ "epoch": 80.9,
1292
+ "learning_rate": 1.597625698324022e-05,
1293
+ "loss": 2.2138,
1294
+ "step": 16100
1295
+ },
1296
+ {
1297
+ "epoch": 81.41,
1298
+ "learning_rate": 1.55572625698324e-05,
1299
+ "loss": 2.1853,
1300
+ "step": 16200
1301
+ },
1302
+ {
1303
+ "epoch": 81.91,
1304
+ "learning_rate": 1.5138268156424581e-05,
1305
+ "loss": 2.2029,
1306
+ "step": 16300
1307
+ },
1308
+ {
1309
+ "epoch": 82.41,
1310
+ "learning_rate": 1.471927374301676e-05,
1311
+ "loss": 2.1583,
1312
+ "step": 16400
1313
+ },
1314
+ {
1315
+ "epoch": 82.91,
1316
+ "learning_rate": 1.4300279329608936e-05,
1317
+ "loss": 2.1613,
1318
+ "step": 16500
1319
+ },
1320
+ {
1321
+ "epoch": 82.91,
1322
+ "eval_cer": 0.30506882326512275,
1323
+ "eval_loss": 1.2093260288238525,
1324
+ "eval_runtime": 123.6333,
1325
+ "eval_samples_per_second": 23.926,
1326
+ "eval_steps_per_second": 2.993,
1327
+ "eval_wer": 0.8718728870858689,
1328
+ "step": 16500
1329
+ },
1330
+ {
1331
+ "epoch": 83.42,
1332
+ "learning_rate": 1.3881284916201116e-05,
1333
+ "loss": 2.1637,
1334
+ "step": 16600
1335
+ },
1336
+ {
1337
+ "epoch": 83.92,
1338
+ "learning_rate": 1.3462290502793295e-05,
1339
+ "loss": 2.1648,
1340
+ "step": 16700
1341
+ },
1342
+ {
1343
+ "epoch": 84.42,
1344
+ "learning_rate": 1.3043296089385474e-05,
1345
+ "loss": 2.1639,
1346
+ "step": 16800
1347
+ },
1348
+ {
1349
+ "epoch": 84.92,
1350
+ "learning_rate": 1.2624301675977652e-05,
1351
+ "loss": 2.1513,
1352
+ "step": 16900
1353
+ },
1354
+ {
1355
+ "epoch": 85.43,
1356
+ "learning_rate": 1.2205307262569831e-05,
1357
+ "loss": 2.1455,
1358
+ "step": 17000
1359
+ },
1360
+ {
1361
+ "epoch": 85.43,
1362
+ "eval_cer": 0.30053506514724504,
1363
+ "eval_loss": 1.2055062055587769,
1364
+ "eval_runtime": 124.9017,
1365
+ "eval_samples_per_second": 23.683,
1366
+ "eval_steps_per_second": 2.962,
1367
+ "eval_wer": 0.8624070317782285,
1368
+ "step": 17000
1369
+ },
1370
+ {
1371
+ "epoch": 85.93,
1372
+ "learning_rate": 1.178631284916201e-05,
1373
+ "loss": 2.1356,
1374
+ "step": 17100
1375
+ },
1376
+ {
1377
+ "epoch": 86.43,
1378
+ "learning_rate": 1.1367318435754188e-05,
1379
+ "loss": 2.1369,
1380
+ "step": 17200
1381
+ },
1382
+ {
1383
+ "epoch": 86.93,
1384
+ "learning_rate": 1.0948324022346368e-05,
1385
+ "loss": 2.1328,
1386
+ "step": 17300
1387
+ },
1388
+ {
1389
+ "epoch": 87.44,
1390
+ "learning_rate": 1.0529329608938546e-05,
1391
+ "loss": 2.1225,
1392
+ "step": 17400
1393
+ },
1394
+ {
1395
+ "epoch": 87.94,
1396
+ "learning_rate": 1.0110335195530725e-05,
1397
+ "loss": 2.1193,
1398
+ "step": 17500
1399
+ },
1400
+ {
1401
+ "epoch": 87.94,
1402
+ "eval_cer": 0.29824776375444184,
1403
+ "eval_loss": 1.1974669694900513,
1404
+ "eval_runtime": 123.9455,
1405
+ "eval_samples_per_second": 23.865,
1406
+ "eval_steps_per_second": 2.985,
1407
+ "eval_wer": 0.8600405679513184,
1408
+ "step": 17500
1409
+ },
1410
+ {
1411
+ "epoch": 88.44,
1412
+ "learning_rate": 9.691340782122904e-06,
1413
+ "loss": 2.1388,
1414
+ "step": 17600
1415
+ },
1416
+ {
1417
+ "epoch": 88.94,
1418
+ "learning_rate": 9.276536312849161e-06,
1419
+ "loss": 2.0962,
1420
+ "step": 17700
1421
+ },
1422
+ {
1423
+ "epoch": 89.45,
1424
+ "learning_rate": 8.85754189944134e-06,
1425
+ "loss": 2.1021,
1426
+ "step": 17800
1427
+ },
1428
+ {
1429
+ "epoch": 89.95,
1430
+ "learning_rate": 8.438547486033518e-06,
1431
+ "loss": 2.0829,
1432
+ "step": 17900
1433
+ },
1434
+ {
1435
+ "epoch": 90.45,
1436
+ "learning_rate": 8.019553072625698e-06,
1437
+ "loss": 2.0911,
1438
+ "step": 18000
1439
+ },
1440
+ {
1441
+ "epoch": 90.45,
1442
+ "eval_cer": 0.30028999714087323,
1443
+ "eval_loss": 1.1960209608078003,
1444
+ "eval_runtime": 122.0182,
1445
+ "eval_samples_per_second": 24.242,
1446
+ "eval_steps_per_second": 3.032,
1447
+ "eval_wer": 0.8647734956051386,
1448
+ "step": 18000
1449
+ },
1450
+ {
1451
+ "epoch": 90.95,
1452
+ "learning_rate": 7.600558659217876e-06,
1453
+ "loss": 2.0836,
1454
+ "step": 18100
1455
+ },
1456
+ {
1457
+ "epoch": 91.46,
1458
+ "learning_rate": 7.1815642458100555e-06,
1459
+ "loss": 2.0527,
1460
+ "step": 18200
1461
+ },
1462
+ {
1463
+ "epoch": 91.96,
1464
+ "learning_rate": 6.762569832402233e-06,
1465
+ "loss": 2.0807,
1466
+ "step": 18300
1467
+ },
1468
+ {
1469
+ "epoch": 92.46,
1470
+ "learning_rate": 6.3435754189944126e-06,
1471
+ "loss": 2.0962,
1472
+ "step": 18400
1473
+ },
1474
+ {
1475
+ "epoch": 92.96,
1476
+ "learning_rate": 5.924581005586592e-06,
1477
+ "loss": 2.0884,
1478
+ "step": 18500
1479
+ },
1480
+ {
1481
+ "epoch": 92.96,
1482
+ "eval_cer": 0.2971449577257689,
1483
+ "eval_loss": 1.1871271133422852,
1484
+ "eval_runtime": 124.4024,
1485
+ "eval_samples_per_second": 23.778,
1486
+ "eval_steps_per_second": 2.974,
1487
+ "eval_wer": 0.8637592968221771,
1488
+ "step": 18500
1489
+ },
1490
+ {
1491
+ "epoch": 93.47,
1492
+ "learning_rate": 5.5055865921787705e-06,
1493
+ "loss": 2.0708,
1494
+ "step": 18600
1495
+ },
1496
+ {
1497
+ "epoch": 93.97,
1498
+ "learning_rate": 5.08659217877095e-06,
1499
+ "loss": 2.067,
1500
+ "step": 18700
1501
+ },
1502
+ {
1503
+ "epoch": 94.47,
1504
+ "learning_rate": 4.667597765363128e-06,
1505
+ "loss": 2.0985,
1506
+ "step": 18800
1507
+ },
1508
+ {
1509
+ "epoch": 94.97,
1510
+ "learning_rate": 4.248603351955307e-06,
1511
+ "loss": 2.0655,
1512
+ "step": 18900
1513
+ },
1514
+ {
1515
+ "epoch": 95.48,
1516
+ "learning_rate": 3.829608938547485e-06,
1517
+ "loss": 2.0766,
1518
+ "step": 19000
1519
+ },
1520
+ {
1521
+ "epoch": 95.48,
1522
+ "eval_cer": 0.296695666380754,
1523
+ "eval_loss": 1.1813979148864746,
1524
+ "eval_runtime": 123.6675,
1525
+ "eval_samples_per_second": 23.919,
1526
+ "eval_steps_per_second": 2.992,
1527
+ "eval_wer": 0.8617308992562542,
1528
+ "step": 19000
1529
+ },
1530
+ {
1531
+ "epoch": 95.98,
1532
+ "learning_rate": 3.4106145251396644e-06,
1533
+ "loss": 2.0485,
1534
+ "step": 19100
1535
+ },
1536
+ {
1537
+ "epoch": 96.48,
1538
+ "learning_rate": 2.9916201117318433e-06,
1539
+ "loss": 2.0725,
1540
+ "step": 19200
1541
+ },
1542
+ {
1543
+ "epoch": 96.98,
1544
+ "learning_rate": 2.572625698324022e-06,
1545
+ "loss": 2.047,
1546
+ "step": 19300
1547
+ },
1548
+ {
1549
+ "epoch": 97.49,
1550
+ "learning_rate": 2.153631284916201e-06,
1551
+ "loss": 2.0649,
1552
+ "step": 19400
1553
+ },
1554
+ {
1555
+ "epoch": 97.99,
1556
+ "learning_rate": 1.7346368715083795e-06,
1557
+ "loss": 2.0735,
1558
+ "step": 19500
1559
+ },
1560
+ {
1561
+ "epoch": 97.99,
1562
+ "eval_cer": 0.2968590450516685,
1563
+ "eval_loss": 1.180108904838562,
1564
+ "eval_runtime": 123.9462,
1565
+ "eval_samples_per_second": 23.865,
1566
+ "eval_steps_per_second": 2.985,
1567
+ "eval_wer": 0.8620689655172413,
1568
+ "step": 19500
1569
+ },
1570
+ {
1571
+ "epoch": 98.49,
1572
+ "learning_rate": 1.3156424581005587e-06,
1573
+ "loss": 2.0277,
1574
+ "step": 19600
1575
+ },
1576
+ {
1577
+ "epoch": 98.99,
1578
+ "learning_rate": 8.966480446927373e-07,
1579
+ "loss": 2.0545,
1580
+ "step": 19700
1581
+ },
1582
+ {
1583
+ "epoch": 99.5,
1584
+ "learning_rate": 4.776536312849162e-07,
1585
+ "loss": 2.0502,
1586
+ "step": 19800
1587
+ },
1588
+ {
1589
+ "epoch": 100.0,
1590
+ "learning_rate": 5.865921787709496e-08,
1591
+ "loss": 2.0333,
1592
+ "step": 19900
1593
+ },
1594
+ {
1595
+ "epoch": 100.0,
1596
+ "step": 19900,
1597
+ "total_flos": 6.621453767453566e+19,
1598
+ "train_loss": 5.91964619928868,
1599
+ "train_runtime": 36587.1788,
1600
+ "train_samples_per_second": 17.438,
1601
+ "train_steps_per_second": 0.544
1602
+ }
1603
+ ],
1604
+ "max_steps": 19900,
1605
+ "num_train_epochs": 100,
1606
+ "total_flos": 6.621453767453566e+19,
1607
+ "trial_name": null,
1608
+ "trial_params": null
1609
+ }